Enhance customer experiences with Apollo GraphOS


<p class="wp-block-paragraph">Apollo’s annual conference, and the world&#8217;s largest GraphQL event, returns to San Francisco in October. <a href="http://www.apollosummit.dev">Apollo Summit 2026</a> runs October 6 to 8, 2026, at <a href="https://themidwaysf.com/">The Midway</a>, with sessions and workshops on how a well-governed graph turns into the entry point for agents and applications alike.</p>



<figure class="wp-block-image size-large"><a href="https://apollosummit.dev/"><img loading="lazy" decoding="async" width="1024" height="538" src="https://wp.apollographql.com/wp-content/uploads/2026/07/Speaker-Ann-Social-1200x630-1-1024x538.jpg" alt="" class="wp-image-25647" srcset="https://wp.apollographql.com/wp-content/uploads/2026/07/Speaker-Ann-Social-1200x630-1-1024x538.jpg 1024w, https://wp.apollographql.com/wp-content/uploads/2026/07/Speaker-Ann-Social-1200x630-1-300x158.jpg 300w, https://wp.apollographql.com/wp-content/uploads/2026/07/Speaker-Ann-Social-1200x630-1-768x403.jpg 768w, https://wp.apollographql.com/wp-content/uploads/2026/07/Speaker-Ann-Social-1200x630-1.jpg 1200w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></a></figure>



<p class="wp-block-paragraph">Apollo has always worked on the difficult middle layer between backend systems and the clients that need to access them, from the early Meteor days through GraphQL’s origins as a way to fetch data for web and mobile apps, and now as the layer that connects AI agents to the systems they need to access in order to do real work. If you own a GraphQL implementation, or you are weighing how your API strategy fits into your AI roadmap, this is where the practical patterns get worked out in public.</p>



<h2 class="wp-block-heading"><a href="#meet-the-speakers" aria-hidden="true" class="aal_anchor" id="meet-the-speakers"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Meet the Speakers</strong></h2>



<p class="wp-block-paragraph">Your APIs were built for apps. What happens when agents want in? These five speakers have already shipped an answer, and they’ll show their work. Stay tuned for more speakers to be announced soon.</p>



<ul class="wp-block-list">
<li><a href="https://www.linkedin.com/in/jonathanwondrusch/">Jonathan Wondrusch</a>, Staff Software Engineer at Block, argues your company already has a graph, hiding in Confluence pages and a hundred half-documented REST endpoints. He shows how to build an API schema that ships product to customers and becomes the context layer your agents reason against. The surprise is that it makes the business legible to itself.</li>



<li><a href="https://www.linkedin.com/in/clarice-abreu/">Clarice Abreu</a>, Senior Software Engineer at Brex, opened GraphOS and found 35% of their federated graph was dead code. Rather than spend months cleaning 60+ subgraphs by hand, she built an agent that deletes, validates, and opens the PR. It has removed over 100,000 lines, and not one customer noticed.</li>



<li><a href="https://www.linkedin.com/in/gijs-martens/">Gijs Martens</a>, Platform Engineer, and <a href="https://www.linkedin.com/in/bartvanlakwijk/">Bart van Lakwijk</a>, Manager Web Platform at Coolblue, routed all production traffic through their graph in week one. The cutover was the easy part. Winning over a &#8220;just do it&#8221; business meant treating the graph as a utility, not a religion, and never blocking a team. No mandate. The graph sold itself.</li>



<li><a href="https://www.linkedin.com/in/sgorla/">Sharon Gorla</a>, Lead Engineer at Starbucks, has a rule for what survives a rewrite: business semantics and shared contracts, never implementation patterns. As coding agents rebuild whole systems, the gap people miss isn&#8217;t MCP or agent frameworks. It&#8217;s alignment. Clients come and go. Shared nouns and verbs are the durable layer.</li>
</ul>



<h2 class="wp-block-heading"><a href="#why-this-matters-now" aria-hidden="true" class="aal_anchor" id="why-this-matters-now"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Why this matters now</strong></h2>



<p class="wp-block-paragraph">Organizations are often pulled in opposing directions: deploying agents quickly vs. keeping control over what they can do. Teams that move fast often hand agents broad access and lose visibility. Teams that lock things down slow every project to a crawl. Summit 2026 is built around a different premise: speed and control are not a tradeoff when the graph sits between your agents and your systems.</p>



<p class="wp-block-paragraph">A self-describing graph gives agents declared, scoped access to exactly the tools they need and nothing more. It reduces token costs by handing agents the right context for each action, enforces field-level security, and improves success rates by making your existing APIs discoverable. The sessions show how teams are already doing this in production.</p>



<h2 class="wp-block-heading"><a href="#what-youll-learn" aria-hidden="true" class="aal_anchor" id="what-youll-learn"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What you&#8217;ll learn</strong></h2>



<p class="wp-block-paragraph">The agenda is organized around four themes that map to the decisions facing platform and AI teams right now.</p>



<ul class="wp-block-list">
<li><strong>Agent infrastructure.</strong> Design the governed entry point between agents and enterprise systems. Build a context layer that improves task success rates while keeping costs down.</li>



<li><strong>Governance and security.</strong> Manage agent identity, start from zero-rights defaults, enforce field-level access policies, and handle credentials in a single layer. Give both developers and non-developers a trusted platform to build on.</li>



<li><strong>Agentic API access.</strong> Connect REST APIs, SQL data sources, MCP servers, and third-party systems into your context layer, whether or not they already live in your graph.</li>



<li><strong>GraphQL and GraphOS.</strong> For teams running GraphQL in production today: the latest in Apollo Federation, Apollo Router optimization, schema governance, Apollo Connectors, and the Apollo MCP Server.</li>
</ul>



<h2 class="wp-block-heading"><a href="#who-youll-meet" aria-hidden="true" class="aal_anchor" id="who-youll-meet"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Who you&#8217;ll meet</strong></h2>



<p class="wp-block-paragraph">Summit draws platform engineers, AI infrastructure leads, developers, and technical leaders who are building API infrastructure for both applications and agents. It is also where the broader GraphQL community gathers, including the people who maintain the specifications and tools the industry runs on. You will leave with patterns you can apply and a set of peers solving the same problems.</p>



<h2 class="wp-block-heading"><a href="#what-to-expect" aria-hidden="true" class="aal_anchor" id="what-to-expect"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What to expect</strong></h2>



<p class="wp-block-paragraph">The format balances learning and doing. Expect keynotes, customer talks, hands-on technical workshops, and time to meet the experts and ask your questions. You can hear about how some of the fastest-moving companies are building agents, learn architectural patterns to apply in your own API strategy, and try it out in a workshop, all in just a few days.</p>



<p class="wp-block-paragraph">Here is the shape of the conference:</p>



<ul class="wp-block-list">
<li><strong>Tuesday, October 6.</strong> Arrivals, registration, and a welcome reception from 5 to 7pm.</li>



<li><strong>Wednesday, October 7.</strong> Sessions and an evening event, 9am to 9pm.</li>



<li><strong>Thursday, October 8.</strong> Sessions and departures, 9am to 3pm.</li>
</ul>



<h2 class="wp-block-heading"><a href="#pricing-and-registration" aria-hidden="true" class="aal_anchor" id="pricing-and-registration"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Pricing and registration</strong></h2>



<p class="wp-block-paragraph">Registration is now open. Get your Summit pass for $399, including breakfast, lunch, and unlimited coffee and snacks across the event. Dietary preferences selectable at sign-up.</p>



<p class="wp-block-paragraph">October 6 to 8, 2026 at The Midway, San Francisco.</p>



<p class="wp-block-paragraph"><a href="https://apollosummit.dev/">Register now.</a></p>


Apollo Summit 2026: Turn Your API Platform Into Your AI Platform


<h2 class="wp-block-heading"><a href="#a-recap-from-graphql-summit-2025-on-how-capital-one-applied-a-product-mindset-to-build-an-enterprise-data-marketplace-powered-by-apollo" aria-hidden="true" class="aal_anchor" id="a-recap-from-graphql-summit-2025-on-how-capital-one-applied-a-product-mindset-to-build-an-enterprise-data-marketplace-powered-by-apollo"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>A recap from GraphQL Summit 2025 on how Capital One applied a product mindset to build an enterprise data marketplace powered by Apollo</h2>



<p class="wp-block-paragraph">Large organizations with complex data ecosystems often face a familiar challenge. As systems evolve over time, data becomes harder to find, harder to trust, and harder to use. Teams build APIs independently, integrations multiply, and identifying the authoritative source of data can take significant effort.</p>



<p class="wp-block-paragraph">This was the broader context behind Capital One’s journey.</p>



<p class="wp-block-paragraph"><em>“All data consumers want to do is find data and build experiences for customers,” </em>explained Susan Price, Product Manager at Capital One.<em> “And all producers want to do is publish their data so consumers will find and use it. Without a holistic solution, organizations may spend a lot of time looking for data, struggle with inconsistent integrations, and validate data quality on their own.”</em></p>



<p class="wp-block-paragraph">While Capital One already had a governed data marketplace in place, they saw the opportunity to build on that foundation to improve speed, reuse, and modern orchestration, creating a more discoverable platform that could serve as a single source of truth across the enterprise.&nbsp;</p>



<p class="wp-block-paragraph"><strong>The Federated GraphQL Approach: Orchestration at the Center</strong></p>



<p class="wp-block-paragraph">Capital One started with a concrete pain point: an internal platform needed to orchestrate data from multiple microservices while avoiding confusion over which systems were authoritative, cached, or redundant. The team saw an opportunity to solve not just this single use case, but a recurring pattern every platform across the company would face.</p>



<p class="wp-block-paragraph">To address it, they brought in GraphQL and <a href="https://www.apollographql.com/docs/graphos/schema-design/federated-schemas/federation">Apollo Federation</a> to scale across teams and compose many services into one cohesive graph.</p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“We brought in GraphQL as a querying framework to handle the orchestration,” </em>said Gaurav Singh, Engineering Leader at Capital One.<em> “With Apollo Federation, we built a central framework that applied our company standards and took care of the heavy lifting, so our teams could focus on business functionality instead of figuring out where the data was coming from.”</em></p>
</blockquote>



<p class="wp-block-paragraph">To prove out their approach, the team routed a subset of their production traffic to their new supergraph. Performance dashboards confirmed low latency, high throughput, and zero disruption to live workloads. The evidence-based rollout built confidence across the organization.</p>



<p class="wp-block-paragraph">With <a href="https://www.apollographql.com/docs/graphos/routing">Apollo Router</a> as the foundation, the team standardized APIs, improved discoverability, and gave teams a transparent way to publish, manage, and govern their data. The marketplace now revolves around five core pillars: <strong>Publish, Manage, Find, Use, and Govern.</strong></p>



<ul class="wp-block-list">
<li><strong>Publish</strong> and <strong>Manage</strong> serve data producers, ensuring data is registered, versioned, and traceable.</li>



<li><strong>Find</strong> and <strong>Use</strong> serve data consumers, enabling quick discovery of authoritative datasets and clear documentation.</li>



<li><strong>Govern</strong> spans all of them, embedding quality checks, compliance, and observability into every step.</li>
</ul>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“From the beginning, observability was paramount,” </em>said Jake Yesbeck, Engineering Manager at Capital One. <em>“We wanted to build a glass house so we could see all the water going through the pipes and all the wiring at once. That visibility let us spot issues before they reached our customers and gave everyone confidence in how the system worked.”</em></p>
</blockquote>



<h2 class="wp-block-heading"><a href="#velocity-through-structure" aria-hidden="true" class="aal_anchor" id="velocity-through-structure"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Velocity Through Structure</strong></h2>



<p class="wp-block-paragraph">What began as a solution for one platform evolved into Capital One&#8217;s enterprise standard for data sharing and orchestration. As teams saw the platform solving real problems, adoption grew organically, and the system scaled right alongside it. In recent months, schema changes proposed against their subgraphs quadrupled, and supergraph deployments tripled, yet the system continued to accelerate rather than slow down.</p>



<p class="wp-block-paragraph">The team&#8217;s investment in embedded governance and automation produced measurable efficiency gains. Organizations building similar data marketplaces have the potential to see benefits as impactful as these example gains:</p>



<ul class="wp-block-list">
<li><strong>50% reduction </strong>in schema composition, proposal, and deployment steps</li>



<li><strong>50% reduction </strong>in communication tools needed (chats, documents, tracking systems)</li>



<li><strong>67% reduction </strong>in people required in the approval loop</li>



<li><strong>300% increase </strong>in deployment velocity</li>
</ul>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“When you spend the time building structure and paving clear paths, you don’t slow people down,” </em>said Jake Yesbeck, Engineering Manager at Capital One. <em>“You actually get more speed because everyone knows the safe path to take.”</em></p>
</blockquote>



<h2 class="wp-block-heading"><a href="#iterating-on-real-producer-and-consumer-feedback" aria-hidden="true" class="aal_anchor" id="iterating-on-real-producer-and-consumer-feedback"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Iterating on Real Producer and Consumer Feedback</strong></h2>



<p class="wp-block-paragraph">Capital One&#8217;s product approach meant continuously listening to their data community. They learned that schema generation tooling had become overloaded with unnecessary custom directives, causing customer-facing errors. They learned that producers wanted visibility into when changes would reach QA and production, who was accountable for next steps, and whether reviews were happening within SLAs.</p>



<p class="wp-block-paragraph">Acting on that feedback, the team expanded authorized approvers, introduced schema observability dashboards, and built a “fast track” workflow for high-priority schema changes.&nbsp;</p>



<p class="wp-block-paragraph">By embedding these improvements directly into workflows, the team maintained compliance while enabling speed at scale.</p>



<h2 class="wp-block-heading"><a href="#from-data-chaos-to-strategic-asset" aria-hidden="true" class="aal_anchor" id="from-data-chaos-to-strategic-asset"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>From Data Chaos to Strategic Asset</strong></h2>



<p class="wp-block-paragraph">What started as a solution for one platform’s orchestration problem became Capital One’s standard approach to data access across the enterprise. More platforms joined organically because it solved real pain points in transparent, measurable ways.</p>



<p class="wp-block-paragraph">The architectural lesson is clear: when you standardize data and make it accessible, predictable, intuitive, and observable, you transform a marketplace from just another source into the source of truth.</p>



<p class="wp-block-paragraph">Capital One’s implementation proves that governance does not have to slow you down. When embedded into workflows rather than bolted on afterward, it becomes the mechanism for sustainable acceleration.</p>



<p class="wp-block-paragraph">“You can only achieve scalable, reliable data access when you treat graph as a platform and embed governance in the workflows,” Price concludes. “Your job is to make the right thing the easy thing.”</p>



<h3 class="wp-block-heading"><a href="#lessons-for-every-enterprise" aria-hidden="true" class="aal_anchor" id="lessons-for-every-enterprise"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Lessons for Every Enterprise</strong></h3>



<p class="wp-block-paragraph">Capital One&#8217;s journey offers a practical blueprint:</p>



<ul class="wp-block-list">
<li><strong>Start with a high-value dataset.</strong> Apply a product lens to its lifecycle from creation to consumption.</li>



<li><strong>Talk to producers and consumers.</strong> Identify friction points and bake governance into solutions.</li>



<li><strong>Build for evolution.</strong> Governance and structure should enable change, not resist it.</li>
</ul>



<p class="wp-block-paragraph">By treating the graph as a long-term product rather than a one-time project, Capital One turned what used to be a bottleneck into a built-in advantage. This approach reflects a broader industry shift where governance is no longer a separate process layered on top of data systems but an integral part of how modern enterprises deliver reliability, compliance, and speed at scale</p>



<p class="wp-block-paragraph">With Apollo, Capital One has established a foundation for continued growth. Teams can now build new data products and AI-powered applications on top of a trusted, observable graph that connects the entire enterprise.</p>



<p class="wp-block-paragraph">Watch the full Capital One session from GraphQL Summit 2025 to hear Susan Price, Gaurav Singh, and Jake Yesbeck share their journey in their own words.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="Transforming Capital One’s Data Marketplace with Embedded Governance and a Product Mindset" width="580" height="326" src="https://www.youtube.com/embed/E_RETbpPogw?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>


Transforming Capital One’s Data Marketplace with Embedded Governance and a Product Mindset


<h2 class="wp-block-heading"><a href="#a-recap-from-graphql-summit-2025-on-how-n-able-turned-certification-automation-and-governance-into-a-system-for-scale" aria-hidden="true" class="aal_anchor" id="a-recap-from-graphql-summit-2025-on-how-n-able-turned-certification-automation-and-governance-into-a-system-for-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>A recap from GraphQL Summit 2025 on how N-able turned certification, automation, and governance into a system for scale.</h2>



<p class="wp-block-paragraph">When N-able’s platform team forecasted their first year of Apollo GraphQL adoption, the plan was conservative: one subgraph, ten engineers, and a year to get it right. What they didn’t anticipate was the demand – developers loved it. Twelve months later, they were running 14 subgraphs, with 85 contributors and 300 Apollo Studio users. The challenge wasn’t infrastructure, it was staying in control through 1,300 percent growth.</p>



<p class="wp-block-paragraph">At the center of that transformation is Dhanik Alkegama, Engineering Manager at N-able and leader of the GraphQL platform team. In his <a href="https://youtu.be/0n82t41v5ro?si=XxloWvUeHYvCYBAG">Summit session </a>he shared how his team survived that growth curve, avoided chaos, and built governance that accelerates instead of blocks.</p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“We planned for one subgraph in 12 months. We ended up with fourteen. That wasn’t a scaling problem, it was a survival problem.”</em> &#8211; Dhanik Alkegama, Engineering Manager, N-able</p>
</blockquote>



<h2 class="wp-block-heading"><a href="#from-api-sprawl-to-a-single-api-with-infinite-possibilities" aria-hidden="true" class="aal_anchor" id="from-api-sprawl-to-a-single-api-with-infinite-possibilities"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>From API Sprawl to a Single API with Infinite Possibilities&nbsp;</strong></h2>



<p class="wp-block-paragraph">N-able builds cybersecurity and IT management software for managed service providers overseeing thousands of devices and endpoints. Years of acquisitions left them with fragmented systems and many REST APIs that behaved almost the same but never quite aligned. Customers were frustrated, and the numbers showed why.</p>



<p class="wp-block-paragraph">A simple question like “which devices are offline?” could require more than 100 API calls: authentication, customer lookup, device enumeration, and then an API request for each of tens of thousands of devices. Some customers managed fleets of 30,000 devices, turning that workflow into thousands of API calls. Many wrote custom PowerShell scripts to orchestrate retries, handle errors across multiple endpoints, and filter payloads with 2,000 fields just to extract the one they needed.</p>



<p class="wp-block-paragraph">“Why do I have to make so many API calls just to get one answer?” one customer asked. “Why do I get 2,000 fields when I only need one?”</p>



<p class="wp-block-paragraph">For partners managing infrastructure at scale, this wasn&#8217;t just friction. It was a reason to leave for competitors with cleaner APIs. When customers started asking for more REST endpoints, leadership made a decision that surprised everyone: no more REST. Everything, from product UI to AI integrations, would flow through a single graph.</p>



<p class="wp-block-paragraph">With Apollo GraphQL, those PowerShell scripts requiring 100+ calls collapsed into a single query returning exactly the needed fields. No loops, no client-side orchestration, no overfetching. The vision was clear: one API with infinite possibilities. The question was how to build it without repeating past mistakes.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2026/01/image-1.png" alt="" class="wp-image-24957" srcset="https://wp.apollographql.com/wp-content/uploads/2026/01/image-1.png 960w, https://wp.apollographql.com/wp-content/uploads/2026/01/image-1-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2026/01/image-1-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<h2 class="wp-block-heading"><a href="#the-breakthrough-education-as-infrastructure" aria-hidden="true" class="aal_anchor" id="the-breakthrough-education-as-infrastructure"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Breakthrough: Education as Infrastructure</strong></h2>



<p class="wp-block-paragraph">Dhanik had seen the challenge before. Six years earlier, he worked on N-able’s first graph, an early effort that left experimental schema hanging around for years and slowed adoption. This time, he wanted a clear plan and a shared understanding from the start.</p>



<p class="wp-block-paragraph">He went to leadership with an architectural proposal, a business case and an idea that would become the foundation for scale: <a href="https://www.apollographql.com/tutorials/#certifications">Apollo certification</a> would be mandatory. Every engineer contributing to the supergraph needed Level 2 certification, and every PM or manager needed Level 1.</p>



<p class="wp-block-paragraph">“You wouldn’t let anyone deploy without a CI/CD pipeline. The same should be true for the graph.”</p>



<p class="wp-block-paragraph">Leadership agreed. They understood that helping everyone speak the same language from day one would ease and accelerate the development of a federated graph built by many teams. Within a year, 154 people were certified, including 132 engineers at Level 2. Reviews became faster, proposals more consistent, and new contributors productive on day one.</p>



<p class="wp-block-paragraph">The impact showed up in reviews and proposals. &#8220;When you have the same training, same understanding of GraphQL, as reviewers we can trust the proposals,&#8221; Alkegama notes. &#8220;For the newer contributors, they are productive from day one.&#8221;</p>



<p class="wp-block-paragraph">Education became infrastructure, a built-in layer of trust that scaled with the team. &#8220;Instead of hoping people will figure it out along the way, we frontloaded the knowledge,&#8221; Alkegama says. &#8220;Discipline can be painful, but so is regret.&#8221;</p>



<h2 class="wp-block-heading"><a href="#governance-that-accelerates" aria-hidden="true" class="aal_anchor" id="governance-that-accelerates"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Governance that Accelerates</strong></h2>



<p class="wp-block-paragraph">With the team aligned and certified, the next step was refining how they worked. N-able’s engineers were moving fast and giving clear feedback that the standard proposal flow of development, testing, and then production was slowing iteration. Rather than enforce a rigid process, the platform team listened.</p>



<p class="wp-block-paragraph">They redesigned the workflow around developer feedback. Instead of running proposals through test and staging environments first, contributors proposed schema changes directly against the production variant.A Lambda-based workflow listens for Apollo Studio webhooks, validates each schema change, runs policy checks, and posts results back to Studio as pass, fail, or warning.</p>



<p class="wp-block-paragraph">“The bottleneck wasn’t the engineers’ knowledge, it was the process,” Alkegama explained. “That realization allowed us to manage fourteen subgraphs with just two or three reviewers.”</p>



<p class="wp-block-paragraph">The system runs asynchronously, so developers keep working while checks complete. Two or three reviewers can now safely govern fourteen subgraphs.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2026/01/image.png" alt="" class="wp-image-24955" srcset="https://wp.apollographql.com/wp-content/uploads/2026/01/image.png 960w, https://wp.apollographql.com/wp-content/uploads/2026/01/image-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2026/01/image-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<h2 class="wp-block-heading"><a href="#what-growth-looks-like" aria-hidden="true" class="aal_anchor" id="what-growth-looks-like"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What Growth Looks Like</strong></h2>



<p class="wp-block-paragraph">In just twelve months, N-able’s numbers tell the story:</p>



<ul class="wp-block-list">
<li><strong>14</strong> subgraphs deployed (from a plan for one)</li>



<li><strong>85</strong> active contributors (from an initial team of 10)</li>



<li><strong>300</strong> Apollo Studio users across the organization</li>



<li><strong>1.3 million+</strong> operations running through the unified graph</li>
</ul>



<p class="wp-block-paragraph">For customers, the change is clear. Workflows that once required hundreds of REST calls now take a single query. The platform delivers what N-able calls a “single API with infinite possibilities,” removing the need for client-side orchestration and manual filtering.</p>



<h2 class="wp-block-heading"><a href="#lessons-for-every-platform-team" aria-hidden="true" class="aal_anchor" id="lessons-for-every-platform-team"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Lessons for Every Platform Team</strong></h2>



<p class="wp-block-paragraph">N-able’s experience offers a simple but powerful blueprint for scaling a graph quickly and safely.</p>



<ol class="wp-block-list">
<li><strong>Make education your first dependency.</strong><strong><br></strong>Certification gave every contributor a shared language and framework for building with confidence. Graph literacy is as critical as a CI/CD pipeline.</li>



<li><strong>Automate governance into the workflow.</strong><strong><br></strong>By embedding checks directly into Apollo Studio, governance became self-service and non-blocking, keeping engineers in flow.</li>



<li><strong>Design for both speed and safety.</strong><strong><br></strong>Production-first proposals allowed N-able to move quickly without sacrificing stability, turning process into a catalyst instead of a constraint.</li>



<li><strong>Build for the graph you will have, not the one you have now.<br></strong>Expect exponential adoption. Systems that can handle ten times your current scale give teams freedom to innovate without fear.</li>
</ol>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“Instead of hoping people will figure it out along the way, we frontloaded the knowledge,” Alkegama says. “Discipline can be painful, but so is regret.”</em></p>
</blockquote>



<h2 class="wp-block-heading"><a href="#what-comes-next" aria-hidden="true" class="aal_anchor" id="what-comes-next"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What Comes Next</strong></h2>



<p class="wp-block-paragraph">N-able is now experimenting with preview contracts and AI-assisted governance to manage experimental features safely. Their next goal is to make the graph the default entry point for all customers, lowering the barrier for teams still using REST APIs.</p>



<p class="wp-block-paragraph">N-able didn’t just scale their graph. They scaled how their teams work together. Watch their full <a href="https://youtu.be/0n82t41v5ro?si=XxloWvUeHYvCYBAG">GraphQL Summit 2025 session</a> to see how they certified to grow, then listen to the <a href="https://www.apollographql.com/events/scaling-apis-smarter-faster-how-n-able-uses-apollo-to-optimize-rapid-growth">Customer Spotlight </a>to hear the story behind it.</p>


How N-able Certified 1300% Growth with Apollo


<h2 class="wp-block-heading"><a href="#at-graphql-summit-2025-delivery-hero-shared-how-a-two-layer-apollo-supergraph-enabled-faster-iteration-and-better-experiences-across-brands-markets-and-devices" aria-hidden="true" class="aal_anchor" id="at-graphql-summit-2025-delivery-hero-shared-how-a-two-layer-apollo-supergraph-enabled-faster-iteration-and-better-experiences-across-brands-markets-and-devices"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>At GraphQL Summit 2025, Delivery Hero shared how a two-layer Apollo supergraph enabled faster iteration and better experiences across brands, markets, and devices.</h2>



<p class="wp-block-paragraph">Delivery Hero’s platform powers food and grocery delivery in more than 70 markets around the globe. When the engineering team set out to speed up UX iteration across its mobile apps, their first attempt, a fully server-driven UI, worked on paper but failed in practice. The custom SDKs were heavy, adoption was slow, and the system meant to accelerate releases became another bottleneck.</p>



<p class="wp-block-paragraph">The breakthrough came when the team split the problem in two: keep business logic in one layer and UI presentation in another. That shift turned months of coordinated releases into a week of configuration work, all powered by Apollo at the core of Delivery Hero’s global supergraph.</p>



<h2 class="wp-block-heading"><a href="#the-cross-platform-iteration-problem" aria-hidden="true" class="aal_anchor" id="the-cross-platform-iteration-problem"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The cross-platform iteration problem</strong></h2>



<p class="wp-block-paragraph">Delivery Hero operates multiple brands, including Foodpanda, Yemeksepeti, and Foodora, on a single platform powered by a global graph that processes billions of queries across its brands each year. In markets where up to 90 percent of orders come from mobile apps, long release cycles were slowing innovation.</p>



<p class="wp-block-paragraph">Product managers wanted the freedom to tweak and experiment with UI without waiting for app releases. Client teams faced duplication and drift across platforms, maintaining separate layout and formatting logic that slowed launches and made consistency hard to achieve.</p>



<p class="wp-block-paragraph">Behind the scenes, restaurant and menu data lived in multiple aggregator services and backends-for-frontends. Each screen re-aggregated similar data, adding complexity and slowing iteration across the stack.</p>



<p class="wp-block-paragraph">The engineering cost was clear, but the strategic cost was greater: slow experimentation in markets where competitors moved faster. In retail and delivery, the ability to test and iterate across channels is not a convenience. It is a necessity.</p>



<figure class="wp-block-pullquote"><blockquote><p>“The goal was to move faster and let product teams run experiments without engineering having to release new apps.” – Arne Wieding, Principal Software Engineer at Delivery Hero</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#the-first-attempt-when-server-driven-goes-too-far" aria-hidden="true" class="aal_anchor" id="the-first-attempt-when-server-driven-goes-too-far"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The first attempt: when server-driven goes too far</strong></h2>



<p class="wp-block-paragraph">Before adopting Apollo, Delivery Hero built <em>Fluid</em>, a complete server-driven UI platform. Clients became “dumb rendering machines,” interpreting backend configurations that controlled everything down to padding and font sizes.</p>



<p class="wp-block-paragraph">Product managers could modify screens instantly, but engineers faced a heavy technical load. “It was too heavy,” said Wieding. The platform collapsed data, UI logic, and configuration into one layer. It required substantial SDKs and created new bottlenecks whenever a team needed a new component.</p>



<p class="wp-block-paragraph">“Fluid taught us what we didn’t want,” Wieding explained: “a too-generic SDUI DSL with a high SDK tax.”</p>



<h2 class="wp-block-heading"><a href="#how-apollo-enabled-delivery-heros-two-layer-architecture" aria-hidden="true" class="aal_anchor" id="how-apollo-enabled-delivery-heros-two-layer-architecture"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>How Apollo enabled Delivery Hero’s two-layer architecture</strong></h2>



<p class="wp-block-paragraph">After retiring their REST-based SDUI system, Delivery Hero rebuilt its platform on Apollo GraphQL, designing the new architecture with Apollo Federation at its core.</p>



<p class="wp-block-paragraph">“We designed everything for Federation from the beginning,” said Wieding. “We didn’t have a monograph”</p>



<p class="wp-block-paragraph">With <a href="https://www.apollographql.com/docs/graphos/routing">Apollo Router</a> orchestrating a single supergraph for Android, iOS, and web, the platform gained a schema-driven foundation that could scale across brands and markets while supporting faster experimentation.</p>



<p class="wp-block-paragraph">The new architecture centered on two “layers” in one supergraph:</p>



<ul class="wp-block-list">
<li><strong>UI layer:</strong> subgraphs that define UI-specific types that return render-ready data and describe appearance without computing logic.</li>



<li><strong>Domain layer:</strong> subgraphs that own canonical business data such as restaurants, products, discounts, and customers.</li>
</ul>



<p class="wp-block-paragraph">Router composes both layers using the <a href="https://www.apollographql.com/docs/graphos/routing/query-planning/query-planning-best-practices#requires-directive">@requires directive</a>. The UI layer declares its dependencies on domain data, and Router fetches everything needed in about 40 milliseconds per vendor tile. “The UI layer just describes; it does not compute,” Wieding explained.</p>



<p class="wp-block-paragraph">Router’s performance makes this model practical, resolving billions of queries across subgraphs while maintaining sub-50 millisecond response times, even under peak load.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2025/11/image-1.png" alt="" class="wp-image-24878" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/image-1.png 960w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-1-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-1-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<h2 class="wp-block-heading"><a href="#focus-on-what-changes-the-tile-strategy" aria-hidden="true" class="aal_anchor" id="focus-on-what-changes-the-tile-strategy"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Focus on what changes: the </strong><strong><em>tile</em></strong><strong> strategy</strong></h2>



<p class="wp-block-paragraph">Instead of making every element configurable, the team focused on what changes most: tiles, the restaurant and product cards users scroll through.</p>



<p class="wp-block-paragraph">Each tile follows a shared schema connected to Delivery Hero’s design system. It combines media, vendor info, incentives, and metadata, all configurable by product managers without code or new releases.</p>



<p class="wp-block-paragraph">When a client requests vendor data via a query, Apollo Router pulls the right information from across the system and returns a finished version of each tile ready for the app to render.</p>



<p class="wp-block-paragraph">“We only need these four fields that is just an array,” said Oluwapelumi Olaoye, Senior Software Engineer at Delivery Hero. Previously, clients requested 80 fields and performed local computations. Now data maps directly to the design system, improving render time by about 100 milliseconds and resolving each tile in roughly 40 milliseconds.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2025/11/image-2.png" alt="" class="wp-image-24880" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/image-2.png 960w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-2-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-2-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<h2 class="wp-block-heading"><a href="#from-three-months-to-one-week" aria-hidden="true" class="aal_anchor" id="from-three-months-to-one-week"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>From three months to one week</strong></h2>



<p class="wp-block-paragraph">The outcomes were clear and measurable:</p>



<ul class="wp-block-list">
<li><strong>Design Cycle Time (3 months → 1 week):</strong> Coordinating design updates across restaurant, cart, and checkout pages now happens in days instead of months.</li>



<li><strong>API Payload Size (80+ fields → 4 fields): </strong>Streamlined, render-ready responses mapped to the design system now load 100 ms faster per view.</li>



<li><strong>5-minute propagation:</strong> Configuration changes sync to production apps within minutes, enabling real-time experimentation and daily iteration.</li>



<li><strong>Adoption by choice:</strong> Teams can still query raw domain data when needed, but most have embraced the model because it provides what they need without adding heavy frameworks or proprietary tools.</li>
</ul>



<p class="wp-block-paragraph">Collectively, these gains transformed experimentation from a scheduled release task into an everyday habit, accelerating innovation across Delivery Hero’s brands and markets.</p>



<h2 class="wp-block-heading"><a href="#why-apollo-router-matters-in-delivery-heros-architecture" aria-hidden="true" class="aal_anchor" id="why-apollo-router-matters-in-delivery-heros-architecture"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Why Apollo Router matters in Delivery Hero’s architecture</strong></h2>



<p class="wp-block-paragraph">For Delivery Hero, Apollo Router is the piece that keeps the two-layer design fast and reliable at global scale. It composes queries across dozens of subgraphs, resolving each vendor tile in about 40 milliseconds, even under heavy load.</p>



<p class="wp-block-paragraph">During peak ordering times, that consistency allows teams to keep experimenting safely. Router’s stability and composition logic ensure the graph can evolve without affecting performance, which is why Delivery Hero treats its supergraph as a core product in its platform rather than infrastructure to work around.</p>



<h2 class="wp-block-heading"><a href="#architectural-principles-that-scale" aria-hidden="true" class="aal_anchor" id="architectural-principles-that-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Architectural principles that scale</strong></h2>



<p class="wp-block-paragraph">Delivery Hero documented its approach in the <a href="https://docs.google.com/document/d/1I8ijaxwsquze8_-1K3hvztF-YD_lmlMIQA7esJr1Zq0/edit?usp=sharing">Supergraph SDUI Guidelines</a>, a shared framework that outlines best practices for schema design, governance, and UI layering. The core idea is simple: keep logic and presentation separate, design for resilience, and give teams flexibility within a common structure.</p>



<p class="wp-block-paragraph">The takeaway from Delivery Hero’s journey is that server-driven UI works when it reduces coordination overhead without replacing engineering judgment. Apollo GraphQL provides the foundation, Apollo Federation powers the supergraph, and Apollo Router delivers the orchestration performance. Together, they enable the experimentation velocity modern retail demands.</p>



<p class="wp-block-paragraph">Watch how Delivery Hero’s platform team brought this two-layer Apollo GraphQL architecture to life at GraphQL Summit 2025. <br></p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="Delivering Faster Cross-Platform UX Iteration at Delivery Hero with Server-Driven Tiles" width="580" height="326" src="https://www.youtube.com/embed/P8MLnGe7Kdo?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>



<p class="wp-block-paragraph"></p>


How Delivery Hero Accelerates UX Experiments with Server-Driven UI and Apollo


<h2 class="wp-block-heading"><a href="#a-recap-from-blocks-graphql-summit-2025-session-on-scaling-graph-adoption-with-apollo-connectors-apollo-mcp-and-ai" aria-hidden="true" class="aal_anchor" id="a-recap-from-blocks-graphql-summit-2025-session-on-scaling-graph-adoption-with-apollo-connectors-apollo-mcp-and-ai"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>A recap from Block’s GraphQL Summit 2025 session on scaling graph adoption with Apollo Connectors, Apollo MCP, and AI</h2>



<p class="wp-block-paragraph">When Block’s GraphQL platform team discovered <a href="https://www.apollographql.com/graphos/apollo-connectors?utm_term=apollo%20connector&amp;utm_campaign=Google_Search_Brand_ApolloGraphQL&amp;utm_source=google&amp;utm_medium=cpc&amp;hsa_acc=6083416509&amp;hsa_cam=20595484542&amp;hsa_grp=173217050114&amp;hsa_ad=720509297031&amp;hsa_src=g&amp;hsa_tgt=kwd-2372618437622&amp;hsa_kw=apollo%20connector&amp;hsa_mt=e&amp;hsa_net=adwords&amp;hsa_ver=3&amp;gad_source=1&amp;gad_campaignid=20595484542&amp;gbraid=0AAAAABiDybZ05Cofg_KP-qsYRLQcuFIHd&amp;gclid=Cj0KCQjw9obIBhCAARIsAGHm1mRuBffxfq57JSMDMxQq4XpR6UD2C3T-QFfNKfRNGbKzMXMGobte74oaAmhSEALw_wcB">Apollo Connectors</a>, they saw a path to bring thousands of existing APIs onto the graph without asking hundreds of engineering teams to run new services or learn new tooling. While Apollo Connectors are typically used for REST APIs, Block found a creative way to connect their gRPC services by leveraging their existing JSON translation layer. Paired with <a href="https://block.xyz/open-source">Goose</a>, Block’s open-source AI agent, and <a href="https://www.apollographql.com/docs/apollo-mcp-server">Apollo MCP Server,</a> that idea became a repeatable system any large organization can use to give developers the speed and flexibility of building on a unified graph.</p>



<figure class="wp-block-pullquote"><blockquote><p>“We started asking the most ambitious question that we could: What if we could use Connectors to bring all of the APIs at Block to our graph?” &#8211; Jonathan Wondrusch, Staff Software Engineer at Block</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#700-teams-3000-services-one-graph" aria-hidden="true" class="aal_anchor" id="700-teams-3000-services-one-graph"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>700 Teams, 3,000 Services, One Graph</strong></h2>



<p class="wp-block-paragraph">Block runs more than 3,000 internal services across 700 teams, all using gRPC and protobuf. Getting a single service onto the federated graph meant learning new tools, deploying a subgraph, and taking on an on-call burden, a cost many teams weren’t willing to take.</p>



<p class="wp-block-paragraph">The business impact was clear. API fragmentation slowed cross-product features, delayed partner integrations, and created barriers to innovation. For a company spanning Square merchants and Cash App customers, every siloed API was a missed opportunity.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2025/11/image.png" alt="" class="wp-image-24858" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/image.png 960w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/image-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<figure class="wp-block-pullquote"><blockquote><p>“Bringing a service into the graph meant learning a new way of working and setting up brand new infrastructure,” said Jonathan Wondrusch, Staff Software Engineer at Block “That’s a learning curve and an engineering cost many teams weren’t willing to take on.”</p></blockquote></figure>



<p class="wp-block-paragraph"><strong>Four Days That Changed Everything</strong></p>



<p class="wp-block-paragraph">Apollo Connectors is designed to connect REST APIs directly to a federated graph, but Block’s services weren’t REST. They used gRPC with protobuf. The breakthrough came when someone noticed their frameworks were already translating protobufs to JSON behind the scenes. That meant they could treat their gRPC services as REST endpoints and use Apollo Connectors without changing backend infrastructure.</p>



<p class="wp-block-paragraph">In a four-day hack week, a dozen teams brought their APIs into the graph with Connectors. No new services. No new infra. The results were so promising that the platform team pivoted its roadmap to make Connectors the default on-ramp to the graph.</p>



<p class="wp-block-paragraph">As adoption grew, the team needed to scale beyond manual schema creation. At first, Goose generated TypeScript subgraphs instead of REST endpoints. After grounding it with Apollo MCP Server and up-to-date documentation, Goose began producing accurate Connector schemas with the right directives, types, and tests.</p>



<h2 class="wp-block-heading"><a href="#making-ai-reliable-at-scale" aria-hidden="true" class="aal_anchor" id="making-ai-reliable-at-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Making AI Reliable at Scale</strong></h2>



<p class="wp-block-paragraph">Rudra Srivastava, Software Engineer at Block, focused on reliability. Raw prompts gave inconsistent results, so the team built Goose Recipes, structured prompt flows that standardize connector generation across teams.</p>



<p class="wp-block-paragraph">The improvement was clear:</p>



<ul class="wp-block-list">
<li><strong>Before</strong>: merchant lists had no pagination and snake_case fields.</li>



<li><strong>After</strong>: Connector schemas used relay-spec pagination, correctly mapped field names, and established entity relationships with inaccessible ID fields.</li>
</ul>



<p class="wp-block-paragraph">Each recipe could be shared across the company, giving every team a clear path to production-ready Connectors.</p>



<figure class="wp-block-pullquote"><blockquote><p>&#8220;Goose recipes make AI behavior consistent and repeatable&#8221; Rudra Srivastava, Software Engineer at Block.</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#from-single-connectors-to-entire-api-surfaces" aria-hidden="true" class="aal_anchor" id="from-single-connectors-to-entire-api-surfaces"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>From Single Connectors to Entire API Surfaces</strong></h2>



<p class="wp-block-paragraph">To go from one Connector to many, the team built Flock, a multi-agent orchestration system that runs multiple Goose agents in parallel. It processes dozens of endpoints, creating fully functional connector subgraphs. The system runs on familiar patterns: queuing, polling, and workers managing four stages through separate queues.</p>



<p class="wp-block-paragraph">A simple sequencing rule prevents type conflicts: if two endpoints modify the same type, agents pick them up in order. This raised success rates and reduced human fixes, while keeping human review where it matters.</p>



<p class="wp-block-paragraph">Block maps authorization at the field level, aligning fields to group-based access and letting <a href="https://www.apollographql.com/docs/graphos/routing">Apollo Router</a> policy prune fields the caller should not see. For migrating existing subgraphs to Connectors, the team uses the <a href="https://www.apollographql.com/docs/graphos/schema-design/federated-schemas/reference/directives">@override directive</a> and its traffic-shift label to cut risk while they move ownership.</p>



<figure class="wp-block-pullquote"><blockquote><p>“Connectors are great for subgraphs that don’t really have business logic on top of their data.” &#8211; Rudra Srivastava, Software Engineer at Block</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#what-changed-in-a-year" aria-hidden="true" class="aal_anchor" id="what-changed-in-a-year"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What Changed in a Year</strong></h2>



<p class="wp-block-paragraph">Block’s journey shows how far “vibe coding” has come, from manual prompts to structured workflows, and from REST APIs to generated user experiences around the graph.</p>



<p class="wp-block-paragraph">They started with one question and ended with a repeatable system:</p>



<ul class="wp-block-list">
<li>Manual prompts became recipes</li>



<li>Recipes evolved into sub-recipes</li>



<li>Sub-recipes scaled through multi-agent orchestration with Flock</li>
</ul>



<p class="wp-block-paragraph">In just a year, what began as an experiment has become a production-ready foundation for how Block builds its platform.</p>



<figure class="wp-block-pullquote"><blockquote><p>“A year ago, we were all wondering if agents, models, and protocols would even work,” the team reflected. “Today, we are heavily invested in building with them, and maybe a year from now, we will wonder how we ever worked without them.”</p></blockquote></figure>



<p class="wp-block-paragraph">To see how they did it, watch  Block’s full GraphQL Summit 2025 session featuring Goose, Flock, and Apollo in action.<br></p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="Going All In on REST Connectors at Block: Vibe coding with Goose" width="580" height="326" src="https://www.youtube.com/embed/_uUmw4xi2sE?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>


How Block Scaled Graph Adoption with Apollo and AI

<h2 class="wp-block-heading"><a href="#at-graphql-summit-wiz-shared-how-the-apollo-mcp-server-helps-bring-security-and-speed-together-in-the-ide" aria-hidden="true" class="aal_anchor" id="at-graphql-summit-wiz-shared-how-the-apollo-mcp-server-helps-bring-security-and-speed-together-in-the-ide"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>At GraphQL Summit, Wiz shared how the Apollo MCP Server helps bring security and speed together in the IDE</h2>



<hr class="wp-block-separator has-alpha-channel-opacity" />



<p class="wp-block-paragraph">Vibe coding, or shipping with AI in the loop, is already here. Developers expect their agents to read the docs, write the call, test it, and move on. Security teams are racing to keep up. At GraphQL Summit 2025, Hen Perez, Field CTO at Wiz, showed how one of the fastest-growing cybersecurity platforms is meeting these realities head-on by integrating MCP and Apollo to keep AI-authored code safe.</p>



<h2 class="wp-block-heading"><a href="#from-manual-steps-to-ai-ready-workflows" aria-hidden="true" class="aal_anchor" id="from-manual-steps-to-ai-ready-workflows"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>From Manual Steps to AI-Ready Workflows</strong></h2>



<p class="wp-block-paragraph">Wiz runs a partner ecosystem with more than 200 certified integrations spanning cloud, identity, and DevOps tools. Wiz’s Integration Network (WIN) follows a clear seven-step process: apply, qualify, sign NDA, access the developer portal, build, certify, and launch joint marketing. Partners across security domains have made it work hundreds of times, but as developer workflows evolve, Wiz saw room to simplify the experience.</p>



<p class="wp-block-paragraph">“[Our current process] is very manual, especially when we think about developers today that are using AI agents as part of their IDEs,” Perez said.</p>



<p class="wp-block-paragraph">Partners were still doing repeated tasks such as authentication setup, query structure, and payload handling that could be standardized and automated. Perez’s goal was clear: bring the entire integration journey into the IDE with validated tools so agents can move quickly and safely while leveraging their use of Apollo GraphQL.</p>



<figure class="wp-block-pullquote"><blockquote><p>“All of our APIs are built with Apollo GraphQL. We have a single supergraph that serves both our customers and partners.” &#8211; Hen Perez, Field CTO at Wiz</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#mcp-turning-the-api-into-the-tool" aria-hidden="true" class="aal_anchor" id="mcp-turning-the-api-into-the-tool"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>MCP: Turning the API Into the Tool</strong></h2>



<p class="wp-block-paragraph">To achieve this, Wiz built on the Model Context Protocol (MCP), a thin layer that lets AI agents talk directly to services instead of translating documentation into code. The solution uses two MCPs together:</p>



<p class="wp-block-paragraph"><strong>Wiz MCP&nbsp;</strong></p>



<ul class="wp-block-list">
<li>Exposes the data from the Wiz platform (issues, vulnerabilities, configurations, inventory).</li>



<li>The WIN API tool surfaces approved, scalable GraphQL queries and integration guidelines inside each tool description.</li>



<li>Agents read what matters and call the right thing.</li>
</ul>



<p class="wp-block-paragraph"><a href="https://www.apollographql.com/docs/apollo-mcp-server"><strong>Apollo MCP Server</strong></a><strong>&nbsp;</strong></p>



<ul class="wp-block-list">
<li>Execute and introspect reveal the actual response structure in the IDE.</li>



<li>Agents verify authentication and wiring against real payloads.</li>
</ul>



<figure class="wp-block-pullquote"><blockquote><p>“All you need to do is combine it with Apollo MCP, which works right out of the box. I haven’t done anything or written a single line of code in the Apollo MCP repository, it just works.” &#8211;&nbsp; Hen Perez, Field CTO at Wiz</p></blockquote></figure>



<p class="wp-block-paragraph">Together they turn API documentation into executable instructions. Agents gain both the validated query (from Wiz) and the real payload shape (from Apollo), providing complete context and removing guesswork.</p>



<p class="wp-block-paragraph">With both MCPs connected, an agent can generate a complete integration from a simple prompt, pulling the WIN API specification, inspecting response structures through Apollo MCP, and producing Python scripts with authentication and data fetching. What once required extensive reading and manual testing now happens in minutes inside the IDE.</p>



<p class="wp-block-paragraph">The generated code follows Wiz’s integration best practices because those guidelines live inside the WIN API tool description. The agent is not improvising; it is using production-ready patterns Wiz has already validated.</p>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="683" src="https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-1024x683.jpg" alt="" class="wp-image-24851" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-1024x683.jpg 1024w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-300x200.jpg 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-768x512.jpg 768w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-1536x1024.jpg 1536w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-2048x1365.jpg 2048w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-1200x800.jpg 1200w, https://wp.apollographql.com/wp-content/uploads/2025/11/TUESDAYFINALS-357-1980x1320.jpg 1980w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<h2 class="wp-block-heading"><a href="#fighting-ai-with-ai" aria-hidden="true" class="aal_anchor" id="fighting-ai-with-ai"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Fighting AI With AI</strong></h2>



<p class="wp-block-paragraph">Perez described a simple reality: developers are already building with AI, and attackers are too. The only viable response is to meet both where they are.</p>



<p class="wp-block-paragraph">“We have to run side by side with developers,” he said. “If the developer is vibe coding, we have to introduce a guardrail. We call it &#8220;vibe security.”</p>



<p class="wp-block-paragraph">Wiz built those guardrails directly into its MCP tools. Security validation happens alongside development, not after it. By grounding AI agents in validated data from Wiz’s platform rather than unpredictable training sources, every suggestion becomes more reliable and secure.</p>



<p class="wp-block-paragraph">When asked about the risk of data poisoning, Perez pointed to MCP descriptions as the safeguard: by defining what the model reads and how it interprets data, Wiz ensures agents operate within trusted boundaries.</p>



<figure class="wp-block-pullquote"><blockquote><p>“Because the attackers are using AI, the developers are using AI, everyone is using AI. In order to really have a successful production environment to be secured and successful, you also have to introduce security within that AI.” &nbsp; Hen Perez, Field CTO at Wiz</p></blockquote></figure>



<h2 class="wp-block-heading"><a href="#the-transferable-pattern" aria-hidden="true" class="aal_anchor" id="the-transferable-pattern"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Transferable Pattern</strong></h2>



<p class="wp-block-paragraph">Wiz’s approach applies to any platform with API-driven integrations.&nbsp;</p>



<ul class="wp-block-list">
<li>Expose a curated, production-safe subset of your API as an MCP tool with clear descriptions and examples.</li>



<li>Pair it with Apollo MCP to let agents execute and introspect live queries, binding code to real payloads instead of mock data.</li>



<li>Put security checks in the same loop so compliance comes with the code.</li>



<li>Evolve the tools with telemetry and add new capabilities without code changes</li>
</ul>



<p class="wp-block-paragraph">As Perez put it, “the beauty here is that you can just take this pattern that we invented and use it for yourself.”</p>



<p class="wp-block-paragraph">Wiz’s approach turns API documentation into a living interface for agents and developers alike. It combines speed, security, and scalability in a single workflow and offers a glimpse of how modern platforms are evolving in the agentic era.</p>



<p class="wp-block-paragraph">You can watch Hen Perez’s full session from GraphQL Summit 2025 and see the demo firsthand.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="How Wiz Is Reimagining Partner Integrations with Apollo MCP" width="580" height="326" src="https://www.youtube.com/embed/RYfuyFNfQ34?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>







<hr class="wp-block-separator has-alpha-channel-opacity" />



<h3 class="wp-block-heading"><a href="#watch-the-playlist-ai-in-dev-workflows-graphql-summit-2025" aria-hidden="true" class="aal_anchor" id="watch-the-playlist-ai-in-dev-workflows-graphql-summit-2025"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Watch the Playlist:</strong><a href="https://youtube.com/playlist?list=PLpi1lPB6opQziSvBbu7kORpqs-SpIzfti&amp;si=b1b6xOiMTXXZ_jhE"><strong> AI in Dev Workflows | GraphQL Summit 2025</strong></a></h3>



<p class="wp-block-paragraph"><em>Talks that show how teams are building agent-ready developer experiences.</em></p>



<p class="wp-block-paragraph">Developers expect their agents to read the docs, write the call, test it, and move on. These sessions from GraphQL Summit 2025 show how leading teams are bringing that reality into their day-to-day workflows.</p>



<hr class="wp-block-separator has-alpha-channel-opacity" />

How Wiz Is Redefining Partner Velocity with Apollo MCP


<h2 class="wp-block-heading"><a href="#insights-shared-live-at-graphql-summit-2025-from-architect-colin-barringer-on-modernizing-healthcare-systems-safely-and-at-scale" aria-hidden="true" class="aal_anchor" id="insights-shared-live-at-graphql-summit-2025-from-architect-colin-barringer-on-modernizing-healthcare-systems-safely-and-at-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Insights shared live at GraphQL Summit 2025 from architect Colin Barringer on modernizing healthcare systems safely and at scale</h2>



<hr class="wp-block-separator has-alpha-channel-opacity" />



<p class="wp-block-paragraph">Clinical software is unforgiving. If the right screen does not update in time, a patient can be harmed. For Athenahealth, that wasn’t theoretical: in one workflow, a nurse could record a latex allergy while a surgeon’s view failed to refresh. Preventing that error became the spark for bringing in Apollo GraphQL.</p>



<p class="wp-block-paragraph">When architect Colin Barringer described the challenge, he didn’t soften the stakes: “If we make the wrong mistake, a person could die.” That level of consequence shaped every decision, from where to start to how quickly they could safely ship.</p>



<p class="wp-block-paragraph">Athenahealth set a clear goal: enable real-time data updates across clinical roles without disrupting a 25-year-old production system used by more than 20,000 medical practices. Achieving that meant threading new technology through legacy code, one verified step at a time.</p>



<h2 class="wp-block-heading"><a href="#the-problem-legacy-code-in-life-or-death-software" aria-hidden="true" class="aal_anchor" id="the-problem-legacy-code-in-life-or-death-software"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Problem: Legacy Code in Life-or-Death Software</strong></h2>



<p class="wp-block-paragraph">Athenahealth’s platform has been evolving since 1997, long before today’s cloud playbooks existed. Over time, thousands of developers contributed code that still ran at scale but carried decades of tradeoffs.</p>



<p class="wp-block-paragraph">One artifact from that history was the data broker, a custom JavaScript library built in 2013 that had become a bottleneck. It efficiently batched data requests in its era, but it didn’t integrate with React, had no dedicated support, and couldn’t stream updates between users. For clinical software, that gap is a risk.</p>



<p class="wp-block-paragraph">The challenge wasn’t just technical. Teams were measured by how fast they shipped features, not how safely they modernized the stack. Everyone agreed modernization was needed, but no one wanted to be the first to slow delivery in an environment where lives and uptime were on the line.</p>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="683" src="https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-1024x683.jpg" alt="" class="wp-image-24837" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-1024x683.jpg 1024w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-300x200.jpg 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-768x512.jpg 768w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-1536x1024.jpg 1536w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-2048x1365.jpg 2048w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-1200x800.jpg 1200w, https://wp.apollographql.com/wp-content/uploads/2025/11/WEDNESDAYFINALS-127-2-1980x1320.jpg 1980w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<h2 class="wp-block-heading"><a href="#the-approach-a-steel-thread-through-legacy-code" aria-hidden="true" class="aal_anchor" id="the-approach-a-steel-thread-through-legacy-code"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Approach: A Steel Thread Through Legacy Code</strong></h2>



<p class="wp-block-paragraph">Barringer borrowed an idea from bridge building: start with a single steel thread across the gap and strengthen it step by step. In software, that means shipping one thin, end-to-end slice in production to prove value and safety before expanding.</p>



<p class="wp-block-paragraph">The team chose the ambulatory surgical center (ASC) workflow as their test case. It required instant data synchronization that the existing stack couldn’t support, but Barringer knew that Apollo’s GraphQL subscriptions could. The specialty also covered just ten practices out of more than 20,000, making it low risk yet high visibility.</p>



<p class="wp-block-paragraph">They changed only what was necessary. The allergy entry page stayed as it was, still using the data broker and proprietary APIs. They added a React micro-frontend with <a href="https://www.apollographql.com/docs/react">Apollo Client</a> connected to <a href="https://www.apollographql.com/docs/apollo-server">Apollo Server</a> running in AWS. The server called Athenahealth’s public REST APIs and listened to a database event queue. When relevant events appeared, it filtered and pushed updates to clients through <a href="https://www.apollographql.com/docs/react/data/subscriptions">Apollo subscriptions</a>.</p>



<p class="wp-block-paragraph">One design choice set the direction for everything that followed. The team built resolvers on Athenahealth’s public APIs instead of internal endpoints, so employees and partners worked from the same surface. Barringer called it “drinking our own champagne.”</p>



<h2 class="wp-block-heading"><a href="#the-results-real-time-safety-and-rapid-adoption" aria-hidden="true" class="aal_anchor" id="the-results-real-time-safety-and-rapid-adoption"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Results: Real-Time Safety and Rapid Adoption</strong></h2>



<p class="wp-block-paragraph">The turning point came during the first live test. A nurse added a latex allergy, and both browsers updated in under one second against a five-second safety budget. Average operation time was ~230 ms, with 30–40 ms from the extra network hop. It was a small moment, but it proved everything: safety, speed, and trust in the new system.</p>



<p class="wp-block-paragraph">Operation volume grew from 200,000 per month to 75 million in seven months. When Apollo reached a mainstream feature, traffic doubled again, reaching ~250 million operations within a year. A single steel thread had become the foundation for real-time care at scale.</p>



<h2 class="wp-block-heading"><a href="#the-platform-infrastructure-as-product" aria-hidden="true" class="aal_anchor" id="the-platform-infrastructure-as-product"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Platform: Infrastructure as Product</strong></h2>



<p class="wp-block-paragraph">Athenahealth treated its platform like a product. The team prioritized consistency and developer experience, delivering templates and tools that let other teams move fast without sacrificing safety.</p>



<p class="wp-block-paragraph">Today, multiple React micro-frontends share a single Apollo Client instance that batches requests, deduplicates operations, and keeps views in sync. Apollo Router federates across three subgraphs: electronic health record data, billing, and event processing, with plans to grow to about twenty as adoption expands.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="960" height="540" src="https://wp.apollographql.com/wp-content/uploads/2025/11/S16-Athenahealth-Clinical-Workflows-Colin-Barringer.pptx.png" alt="" class="wp-image-24835" srcset="https://wp.apollographql.com/wp-content/uploads/2025/11/S16-Athenahealth-Clinical-Workflows-Colin-Barringer.pptx.png 960w, https://wp.apollographql.com/wp-content/uploads/2025/11/S16-Athenahealth-Clinical-Workflows-Colin-Barringer.pptx-300x169.png 300w, https://wp.apollographql.com/wp-content/uploads/2025/11/S16-Athenahealth-Clinical-Workflows-Colin-Barringer.pptx-768x432.png 768w" sizes="auto, (max-width: 960px) 100vw, 960px" /></figure>



<h2 class="wp-block-heading"><a href="#the-next-frontier-connecting-ai-agents-safely" aria-hidden="true" class="aal_anchor" id="the-next-frontier-connecting-ai-agents-safely"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Next Frontier: Connecting AI Agents Safely</strong></h2>



<p class="wp-block-paragraph">Athenahealth is also experimenting with <a href="https://www.apollographql.com/docs/apollo-mcp-server">Apollo MCP Server</a> to connect AI agents safely to their supergraph. In a live demo, an AI assistant used the graph to retrieve insights about social determinants of health directly from Apollo GraphQL. Setup was simple: they whitelisted safe queries and let the agent use those tools.</p>



<p class="wp-block-paragraph">Barringer emphasized that trust and safety remain paramount as they design controls to prevent hallucinations or unsafe data exposure. The company’s legal and engineering teams are now collaborating on frameworks to define and validate trustworthy agent behavior.</p>



<h2 class="wp-block-heading"><a href="#lessons-you-can-apply" aria-hidden="true" class="aal_anchor" id="lessons-you-can-apply"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Lessons You Can Apply</strong></h2>



<p class="wp-block-paragraph">Athenahealth’s experience offers a repeatable pattern for organizations modernizing legacy systems in regulated, high-stakes environments:</p>



<ul class="wp-block-list">
<li>Identify a use case that cannot be solved with existing technology.</li>



<li>Start with a small, controlled scope but real business value</li>



<li>Deploy end-to-end in production to prove value and safety.</li>



<li>Document decisions and scale with governance, not guesswork</li>



<li>Invest early in shared tooling and templates to sustain safe velocity</li>
</ul>



<p class="wp-block-paragraph">As Barringer summarized, “The steel thread approach allows us to build the bridge to new tech in midair.”</p>



<p class="wp-block-paragraph">For teams considering Apollo GraphQL in mission-critical systems, Athenahealth’s story shows that careful constraints like patient safety, legacy preservation, and limited scope can become advantages. Once developers see real value and ease of use, adoption accelerates on its own.</p>



<p class="wp-block-paragraph">Watch Colin Barringer’s full session and demo from GraphQL Summit 2025.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="High Stakes, Real Time: How Athenahealth Scaled GraphQL in Clinical Workflows" width="580" height="326" src="https://www.youtube.com/embed/g42416YYuBw?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>


When Every Second Matters: How Athenahealth Protects Patients with Apollo GraphQL

Building Trust into AI: How GraphQL Architects Accelerated Developer Delivery

10 Years of GraphQL Celebrated at GraphQLConf 2025


<p class="wp-block-paragraph"><kbd>This post is part of our GraphQL Summit 2024 customer series, highlighting how leading teams are using Apollo to accelerate development and build world-class applications. Join hundreds of your peers this October as we return for <a href="https://www.apollographql.com/graphql-summit-2025">GraphQL Summit 2025.</a></kbd></p>



<hr class="wp-block-separator has-alpha-channel-opacity" />



<p class="wp-block-paragraph">&#8220;Like all great software companies, we built it twice,&#8221; Mike Cohen shared with the GraphQL Summit audience, pausing for laughter. As Technical Fellow at Indeed, he was describing a decision that looked like organizational dysfunction turned into strategic advantage: running two competing GraphQL platforms simultaneously before unifying them into OneGraph, which now processes <strong>100 billion requests per month</strong> across <strong>295 subgraphs</strong>.</p>



<p class="wp-block-paragraph">For business leaders wrestling with API strategy decisions, Indeed&#8217;s journey offers a counterintuitive lesson about when duplication becomes discovery.</p>



<h2 class="wp-block-heading"><a href="#the-cost-of-api-chaos" aria-hidden="true" class="aal_anchor" id="the-cost-of-api-chaos"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Cost of API Chaos</strong></h2>



<p class="wp-block-paragraph">Five years ago, Indeed&#8217;s API landscape was costing them market opportunities. With no central strategy, hundreds of APIs sprawled across teams with inconsistent protocols and security practices. Developers couldn&#8217;t find existing APIs, forcing them to &#8220;ask around and hope to find a team that would give you access,&#8221; as Cohen put it.</p>



<p class="wp-block-paragraph">The business impact was tangible: engineering velocity slowed, critical partner integrations like Glassdoor faced delays, and new market initiatives stalled on integration complexity. For a company helping millions find jobs while serving employers and hundreds of partner organizations, API fragmentation was becoming a tax on growth.</p>



<p class="wp-block-paragraph">Nina Mutty, Senior Technical Product Manager at Indeed, saw this firsthand in the employer-focused products: &#8220;Teams were building APIs in silos with limited reuse across the organization. It was hard to scale new initiatives efficiently.&#8221; Each team’s distinct approach was furthering silos.</p>



<h2 class="wp-block-heading"><a href="#the-parallel-platform-experiment" aria-hidden="true" class="aal_anchor" id="the-parallel-platform-experiment"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>The Parallel Platform Experiment</strong></h2>



<p class="wp-block-paragraph">But instead of forcing a single solution, Indeed&#8217;s leadership made an unconventional choice: fund API platform teams in two business units to solve the same problem differently and learn which might offer more for the company. The employer team built EmployerGraph using schema stitching, while the jobseeker team created Jobseeker Federator using Apollo Federation.</p>



<p class="wp-block-paragraph">&#8220;The Jobseeker team reached out to the Employer team and tried to get them to combine forces because duplication seemed bad… but neither solution was proven,&#8221; Mutty recalls. &#8220;It was just untenable at the time.&#8221;</p>



<p class="wp-block-paragraph">This parallel approach let Indeed test two technical strategies simultaneously while teams learned what worked in their specific contexts. Both platforms shared the same vision: centralized discovery, consistent standards, reusability, and strong developer experience.&nbsp;</p>



<p class="wp-block-paragraph">Early wins from both the GraphQL implementations validated the overall strategy. API developers could reuse capabilities while maintaining autonomy and discovery problems disappeared. Most importantly for Indeed&#8217;s micro frontend architecture, data sharing became dramatically simpler, eliminating the &#8220;manually wrangling code and redux stores&#8221; that had plagued frontend development.</p>



<h2 class="wp-block-heading"><a href="#choosing-the-winner" aria-hidden="true" class="aal_anchor" id="choosing-the-winner"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Choosing the Winner</strong></h2>



<p class="wp-block-paragraph">By fall 2021, the data was clear. Jobseeker Federator had proven itself at scale, serving <strong>3.5 billion requests per month</strong> and successfully onboarding Glassdoor as Indeed&#8217;s first major external customer. Apollo Federation&#8217;s schema registry provided better safety and validation, while its decentralized approach offered more flexibility for extending types across teams.</p>



<p class="wp-block-paragraph">The employer organization made the strategic call: abandon EmployerGraph and migrate to Apollo Federation. This meant shutting down their schema stitching approach and moving all employer APIs to what would become OneGraph, with an aggressive two-year timeline to maintain momentum.</p>



<p class="wp-block-paragraph">&#8220;We really only had about two years to make this migration happen or we&#8217;d risk losing that support and momentum,&#8221; Mutty explained. This urgency shaped their approach: focus on adoption over perfection.</p>



<h2 class="wp-block-heading"><a href="#platform-results-at-scale" aria-hidden="true" class="aal_anchor" id="platform-results-at-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Platform Results at Scale</strong></h2>



<p class="wp-block-paragraph">The migration succeeded, delivering impressive scale:</p>



<ul class="wp-block-list">
<li><strong>100 billion requests per month</strong> globally</li>



<li><strong>295 subgraphs</strong> maintained by <strong>200+ engineering teams</strong></li>



<li><strong>9,000 types</strong> in the unified graph</li>
</ul>



<p class="wp-block-paragraph">This scale was achieved through automated developer workflows that eliminated traditional bottlenecks. Indeed implemented push-button subgraph creation with built-in best practices, automated deployment pipelines, and integrated schema checks and lint rules into every merge request. All subgraphs automatically include observability through Datadog integration, reducing operational overhead as the platform grew.</p>



<p class="wp-block-paragraph">The business impact extended beyond internal efficiency. Indeed successfully integrated Glassdoor for job search and matching capabilities, launched Indeed PLUS in Japan to expose APIs to third-party job boards, and reduced duplication and maintenance overhead across customer platforms. The platform survived and thrived through Indeed&#8217;s recent organizational shift, merging jobseeker and employer units into a unified marketplace. OneGraph already reflected this structure, validating the federated approach.</p>



<h2 class="wp-block-heading"><a href="#platform-challenges-and-evolution" aria-hidden="true" class="aal_anchor" id="platform-challenges-and-evolution"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Platform Challenges and Evolution</strong></h2>



<p class="wp-block-paragraph">The migration succeeded, but success brought new complexities. Traffic surged from <strong>5 billion to 100 billion requests per month</strong>. The core platform team faced what Cohen calls &#8220;success overload,&#8221; buried in requests as hundreds of teams onboarded.</p>



<p class="wp-block-paragraph">Despite establishing GraphQL guilds, automated tooling, and formal API reviews, the rapid pace of adoption created new challenges that would need to be addressed as the platform matured.</p>



<h2 class="wp-block-heading"><a href="#strategic-investment-in-partner-experience" aria-hidden="true" class="aal_anchor" id="strategic-investment-in-partner-experience"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Strategic Investment in Partner Experience</strong></h2>



<p class="wp-block-paragraph">Indeed is now leveraging OneGraph&#8217;s proven scalability for strategic business expansion. Following the successful Japan launch, they&#8217;re investing heavily in partner capabilities, addressing challenges that emerged from rapid platform growth:&nbsp;</p>



<p class="wp-block-paragraph"><strong>Domain Model Refinement</strong>: Updating internal concepts to align with industry-standard terminology, reducing friction for external partners while addressing some of the inconsistencies that developed during the rapid scaling phase.</p>



<p class="wp-block-paragraph"><strong>Graph Organization</strong>: With over 9,000 types in the graph, Indeed has “sort of reintroduced a discoverability problem,” as Mutty noted. “Users really only care about a specific slice of the graph.” They’re exploring namespacing and experience-based variants to serve different user needs &#8211; job search and matching (sourcing) versus job posting and advertising (management)</p>



<p class="wp-block-paragraph">The company is also implementing differentiated governance standards: flexible for internal experimentation, elevated quality for external production APIs. As Cohen reflects, while their GraphQL guilds and automated tooling helped during the migration, “they might not have been quite strong enough” for the scale they had reached.</p>



<h2 class="wp-block-heading"><a href="#lessons-for-api-strategy-leaders" aria-hidden="true" class="aal_anchor" id="lessons-for-api-strategy-leaders"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Lessons for API Strategy Leaders</strong></h2>



<figure class="wp-block-pullquote"><blockquote><p><em>&#8220;Don&#8217;t underestimate what you can get done with stewardship, but recognize when you need stronger governance. Our graph wouldn&#8217;t be where it is today without very active stewards who care about the quality of the graph.&#8221;</em> <em>— Mike Cohen, Technical Fellow at Indeed</em></p></blockquote></figure>



<p class="wp-block-paragraph">Cohen&#8217;s insight challenges conventional wisdom about avoiding duplication. Indeed&#8217;s parallel platform approach worked because it generated real data about what mattered: decentralized ownership, proven scalability, external integration capabilities. The &#8220;failed&#8221; EmployerGraph wasn&#8217;t a waste &#8211; it was research that informed the winning choice.</p>



<p class="wp-block-paragraph">For business leaders considering API platform investments, Indeed&#8217;s journey demonstrates that sometimes the fastest path to the right solution involves building competing approaches first. The key is having clear criteria for choosing the winner and the discipline to act on the data.</p>



<p class="wp-block-paragraph">To learn more from Indeed, watch the full session from GraphQL Summit 2024.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="OneGraph: Indeed as a platform" width="580" height="326" src="https://www.youtube.com/embed/5JMIymo3NWY?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>


How Indeed’s Bold Bet on Parallel API Platforms Paid Off


<p class="wp-block-paragraph"><em>The world&#8217;s largest and longest-running GraphQL event returns to San Francisco with 30+ sessions on building agents, MCP services, and modern API platforms</em></p>



<p class="wp-block-paragraph">Whether you’re building mobile apps, web experiences, or preparing for an agentic future, the challenge is the same: orchestrating your organization’s APIs is far more complex than it appears. And that&#8217;s why leading organizations are turning to GraphQL as their API orchestration layer.</p>



<p class="wp-block-paragraph">At Indeed, they&#8217;re reimagining talent discovery with GraphQL as the central interface between agents to enable a seamless, intelligent hiring experience. Intuit is building smart tools to eliminate bottlenecks, enabling developers to ship their supergraph with speed and confidence. And at Booking.com, they’re continuously modernizing their architecture to stay ahead of their scale, reduce operational overhead and improve performance.</p>



<p class="wp-block-paragraph">And we&#8217;re excited to welcome these industry leaders to the stage to share more at our next <a href="http://www.graphqlsummit.com">GraphQL Summit by Apollo 2025</a>, our flagship event and the 8th annual edition of the conference that started it all, taking place October 6-8, 2025 at the Hyatt Regency San Francisco.</p>



<h1 class="wp-block-heading"><a href="#a-declarative-approach-to-building-connected-applications" aria-hidden="true" class="aal_anchor" id="a-declarative-approach-to-building-connected-applications"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>A Declarative Approach to Building Connected Applications</strong></h1>



<p class="wp-block-paragraph">This year marks a special milestone – it’s been 10 years since the GraphQL specification was first made available to the world. A decade of innovation, adoption, and transformation in how we think about APIs and connected systems.</p>



<p class="wp-block-paragraph">And with GraphQL emerging as the API orchestration layer that makes today and tomorrow’s applications possible, this year&#8217;s Summit addresses the questions that matter most: How do you build systems that can truly connect everything while maintaining security and performance? How do you architect APIs that enable teams to build anything, from traditional applications to AI-powered experiences?</p>



<p class="wp-block-paragraph">GraphQL has evolved far beyond a query language. What started by reimagining web and mobile development is now expanding into new frontiers. Engineering teams are exploring how Apollo serves as the API orchestration layer for emerging use cases like AI agents, Model Context Protocol (MCP) services, and interconnected systems.</p>



<h1 class="wp-block-heading"><a href="#industry-leaders-share-real-world-breakthroughs" aria-hidden="true" class="aal_anchor" id="industry-leaders-share-real-world-breakthroughs"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Industry Leaders Share Real-World Breakthroughs</strong></h1>



<p class="wp-block-paragraph">The early speaker lineup demonstrates the breadth and depth of Graph-based API Orchestration innovation happening across industries:</p>



<ul class="wp-block-list">
<li><a href="https://www.linkedin.com/in/christian-ernst11/"><strong>Christian Ernst</strong></a>, Senior Software Engineer at Booking.com, details<strong> &#8220;Inside Booking.com&#8217;s Router Rollout: Powering Product and Platform at Massive Scale,&#8221;</strong> offering insights from one of the world&#8217;s largest travel platforms.</li>



<li><a href="https://www.linkedin.com/in/jeffrey-russom-0a520436/"><strong>Jeffrey Russum</strong></a>, Principal Software Engineer at Indeed, delivers <strong>&#8220;Powering Indeed&#8217;s Agentic Talent Scout with Federated Subscriptions,&#8221;</strong> revealing how the world&#8217;s largest job platform is building AI-powered experiences with GraphQL federation.</li>



<li><strong>Oleks Bidiuk</strong>, Software Engineer, at Intuit, presents <strong>“Stop Reviewing Schemas: How Intuit’s AI Schema Co-Pilot Made Developers Faster by Automating Governance,”</strong> demonstrating how one of the world&#8217;s largest financial platforms eliminates governance bottlenecks at scale.</li>



<li><a href="https://www.linkedin.com/in/wibodha/?originalSubdomain=uk"><strong>Dhanik Alkegama</strong></a>, Engineering Manager &#8211; Platform Services at N-able, shares<strong> &#8220;Certified to Scale: How N-able Hit 1300% GraphQL Growth in 12 Months,&#8221;</strong> providing proven strategies for explosive GraphQL adoption.</li>



<li><a href="https://www.linkedin.com/in/rohitgupta4/"><strong>Rohit Gupta</strong></a>, Platform Product Manager, and <a href="https://www.linkedin.com/in/bhavanapallempati/"><strong>Bhavana Sree Pallempati</strong></a>, Software Engineer, Wayfair, presents <strong>“AI Meets PlatformOps: Auto-Mocking and Documenting Schemas at Wayfair with GenAI,”</strong> showing how GraphQL enables rapid innovation in e-commerce.</li>



<li><a href="https://www.linkedin.com/in/henp/?originalSubdomain=il"><strong>Hen Perez</strong></a>, Field CTO for API Infrastructure at Wiz, presents <strong>&#8220;How Wiz Is Reimagining Partner Integrations with Apollo MCP,&#8221;</strong> demonstrating cutting-edge approaches to partner connectivity in the cybersecurity space.</li>
</ul>



<h1 class="wp-block-heading"><a href="#what-youll-experience" aria-hidden="true" class="aal_anchor" id="what-youll-experience"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>What You’ll Experience</strong></h1>



<p class="wp-block-paragraph">GraphQL Summit 2025 delivers unparalleled learning and networking opportunities:</p>



<ul class="wp-block-list">
<li><strong>Keynotes</strong> featuring industry leaders sharing transformative insights on the future of GraphQL and connected systems.</li>



<li><strong>Breakout Sessions </strong>covering real-world implementations, best practices, and cutting-edge use cases from companies successfully deploying GraphQL at scale.</li>



<li><strong>Meet the Apollo Experts </strong>sessions providing direct access to the team behind Federation and Apollo&#8217;s latest innovations.</li>



<li><strong>Connectors Hackathon </strong>where you can get hands-on experience building with Apollo&#8217;s newest tools and technologies.</li>



<li><strong>AI + Connectors Certification </strong>to validate your expertise in building AI-powered applications with GraphQL.</li>
</ul>



<p class="wp-block-paragraph">Beyond technical content, GraphQL Summit creates opportunities to connect with the broader GraphQL ecosystem. The conference includes a welcome reception, evening event celebration, and plenty of space for networking, connection, and demos throughout the event.</p>



<p class="wp-block-paragraph">Partnership opportunities are available – reach out to <a href="mailto:events@apollographql.com">events@apollographql.com</a> if interested in partnering.</p>



<p class="wp-block-paragraph"><strong>Ready to Connect Everything and Build Anything?</strong></p>



<p class="wp-block-paragraph">Join us in San Francisco at the Hyatt Regency on October 6-8, 2025, to learn from the best, connect with innovators, and discover how GraphQL can unlock new possibilities for your organization.</p>



<p class="wp-block-paragraph"><strong>Early bird registration is now open at $599</strong> with special rates available for Apollo customers and community contributors. Early bird pricing ends August 1, 2025.</p>



<p class="wp-block-paragraph">Register now at <a href="http://www.graphqlsummit.com">www.graphqlsummit.com</a>.</p>



<p class="wp-block-paragraph">Thank you to our community and partners for the privilege of working together and delivering more and more impact with GraphQL, every day.</p>



<p class="wp-block-paragraph">Here&#8217;s to GraphQL Summit 2025!</p>



<figure class="wp-block-embed is-type-video is-provider-wistia-inc wp-block-embed-wistia-inc"><div class="wp-block-embed__wrapper">
<script src="https://fast.wistia.com/player.js" async></script><script src="https://fast.wistia.com/embed/f1u17mz9bs.js" async type="module"></script><style>wistia-player[media-id='f1u17mz9bs']:not(:defined) { background: center / contain no-repeat url('https://fast.wistia.com/embed/medias/f1u17mz9bs/swatch'); display: block; filter: blur(5px); padding-top:56.25%; }</style> <wistia-player media-id="f1u17mz9bs" seo="false" dnt="1" aspect="1.7777777777777777"></wistia-player>
</div></figure>


Announcing GraphQL Summit by Apollo 2025: Connect Everything. Build Anything.


<p class="wp-block-paragraph"><kbd>This post is part of our GraphQL Summit 2024 customer series, highlighting how leading teams are using Apollo to accelerate development and build world-class applications. Join hundreds of your peers this October as we return for <a href="https://www.apollographql.com/graphql-summit-2025">GraphQL Summit 2025.</a></kbd></p>



<hr class="wp-block-separator has-alpha-channel-opacity" />



<p class="wp-block-paragraph"><a href="https://www.booking.com/">Booking.com</a>, a leader in the online travel industry, has significantly advanced its digital infrastructure by integrating <a href="https://www.apollographql.com/docs/graphos/schema-design/federated-schemas/federation">Apollo Federation</a> into its service architecture. At GraphQL Summit 2024, Sanver Tarmur, Senior Software Engineer at Booking.com discussed the company&#8217;s journey to improve user experience with GraphQL and how they transformed their approach to API orchestration – by moving to a distributed, federated model.</p>



<h2 class="wp-block-heading"><a href="#context-and-challenges" aria-hidden="true" class="aal_anchor" id="context-and-challenges"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Context and Challenges</h2>



<p class="wp-block-paragraph">Before Federation, Booking.com faced significant API orchestration challenges. With over 1,000 simultaneous experiments running across their platform and connections to multiple domains (accommodations, flights, attractions, and vehicle rentals), the company&#8217;s monolithic orchestration layer became increasingly complex.</p>



<p class="wp-block-paragraph">“Our monograph setup was essentially a centralized API orchestration layer,&#8221; explained Sanver. &#8220;It was responsible for connecting clients with all our backend services, transforming data, handling errors, and coordinating the complex relationships between our various domains.&#8221;</p>



<p class="wp-block-paragraph">This approach to API orchestration created several issues:</p>



<ul class="wp-block-list">
<li>Performance bottlenecks as all orchestration logic flowed through a single system</li>



<li>Development bottlenecks as teams waited for changes to the central orchestration layer</li>



<li>Difficulty maintaining consistent experiences across different client platforms</li>



<li>Limited ability to optimize orchestration patterns for specific domains</li>
</ul>



<h2 class="wp-block-heading"><a href="#the-shift-to-apollo-federation" aria-hidden="true" class="aal_anchor" id="the-shift-to-apollo-federation"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The Shift to Apollo Federation</h2>



<p class="wp-block-paragraph">Booking.com’s GraphQL journey began with a monograph setup, an initial schema stitching layer primarily focused on its highest-traffic business domain, accommodations. “We sell around one billion reservations per year, that is the most critical part of our business, we wanted to start with that part,&#8221; Sanver shared. The goal was to streamline client communication without embedding business logic into the monograph layer.</p>



<p class="wp-block-paragraph">However, as adoption grew, so did the challenges. “Our initial setup became a bottleneck. Developer experience suffered, performance issues arose, and schema governance became difficult to maintain,” Sanver noted. At its peak, this setup supported 35 subgraphs, but the centralized team managing it became overwhelmed. Performance issues became particularly critical, with Sanver noting that a &#8220;20% latency increase causes a lot of booking loss for us”</p>



<h2 class="wp-block-heading"><a href="#apollo-federation-implementation" aria-hidden="true" class="aal_anchor" id="apollo-federation-implementation"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Apollo Federation Implementation</h2>



<p class="wp-block-paragraph">The migration to Apollo Federation enabled Booking.com to decouple subgraphs, improve performance, and enhance developer experience through improved independent delivery cadence. The monograph became the first subgraph in the federation gateway, gradually giving way to a more distributed architecture. The team was able to decentralize the ownership of GraphQL schemas which significantly enhanced developer experience and reduced schema governance challenges. As Sanver explained, &#8220;we helped people to migrate their schemas out of that single subgraph, interfaced by federation gateway, and we encouraged adaptation internally in the company.&#8221;</p>



<p class="wp-block-paragraph">Later, during their cloud modernization, Booking.com undertook a strategic shift from their internal clusters to AWS. The team implemented edge routing for query optimization, enhanced traffic management using Envoy, and added resilience measures including schema caching and timeout management systems to overcome cross-cloud communication challenges.</p>



<p class="wp-block-paragraph">To ensure reliable operations at scale, Booking.com implemented comprehensive monitoring and governance measures:</p>



<ul class="wp-block-list">
<li>Query-level latency monitoring to track and optimize query performance</li>



<li>Timeout budgets dynamically adjusted based on query plans</li>



<li>Schema caching to prevent issues during schema fetch failures</li>



<li>Comprehensive alerting and monitoring dashboards to maintain reliability</li>
</ul>



<h2 class="wp-block-heading"><a href="#measurable-results" aria-hidden="true" class="aal_anchor" id="measurable-results"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Measurable Results</h2>



<h3 class="wp-block-heading"><a href="#technical-scale" aria-hidden="true" class="aal_anchor" id="technical-scale"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Technical Scale</strong></h3>



<p class="wp-block-paragraph">&#8220;With Federation, we now support 140 service connections handling over 8 billion requests per day, with peaks of 100,000 requests per second,&#8221; Sanver shared. This represents one of the largest API orchestration implementations in the travel industry, coordinating data flows across accommodations, flights, attractions, and vehicle rentals to create seamless experiences for Booking.com&#8217;s 120 million active clients across iOS, Android, and Web.</p>



<h3 class="wp-block-heading"><a href="#development-velocity" aria-hidden="true" class="aal_anchor" id="development-velocity"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Development Velocity</strong></h3>



<p class="wp-block-paragraph">The most significant business impact? Product delivery accelerated by 40%, allowing Booking.com to respond faster to market opportunities and customer needs. As Sanver shared, &#8220;developers don&#8217;t need to wait for release cycles anymore. They can ship quite faster compared to before.”</p>



<h3 class="wp-block-heading"><a href="#internal-efficiency" aria-hidden="true" class="aal_anchor" id="internal-efficiency"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Internal Efficiency</strong></h3>



<p class="wp-block-paragraph">Beyond external benefits, Federation also solved internal challenges. Service-to-service communication, particularly for backend services that needed access to data, became significantly more efficient. &#8220;Instead of every service connecting directly to different databases, we used federation to streamline access, reduce latency, and improve performance,&#8221; Sanver explained.<br></p>



<h3 class="wp-block-heading"><a href="#developer-experience" aria-hidden="true" class="aal_anchor" id="developer-experience"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Developer Experience</strong></h3>



<p class="wp-block-paragraph">Apollo Federation enabled developers to independently manage their services without the bottlenecks typically associated with monolithic architectures. This autonomy has made developers at Booking.com happier and more productive. The company has put in a lot of effort to educate its teams on GraphQL best practices, building internal tools to monitor schema quality, and create a supportive environment for ongoing learning and improvement.</p>



<h2 class="wp-block-heading"><a href="#future-outlook-and-strategic-initiatives" aria-hidden="true" class="aal_anchor" id="future-outlook-and-strategic-initiatives"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Future Outlook and Strategic Initiatives</h2>



<p class="wp-block-paragraph">Booking.com will continue to refine its use of API orchestration approach, with plans to migrate to <a href="https://www.apollographql.com/docs/graphos/routing">Apollo Router</a> to further enhance performance and cost-efficiency. This next generation of orchestration technology promises significant performance improvements: complex orchestration use cases saw latency reduction from 200 milliseconds to 10 milliseconds, while property page interactions (which requires orchestrating hotel data from multiple services) saw overhead decrease from around 100 milliseconds to 5 milliseconds.</p>



<figure class="wp-block-pullquote"><blockquote><p><em>&#8220;As we continue to evolve our API orchestration strategy,&#8221; Sanver noted, &#8220;we&#8217;re seeing not just technical benefits but business impacts as well. More efficient orchestration directly translates to faster page loads, better conversion rates, and ultimately, more bookings.&#8221;</em></p></blockquote></figure>



<p class="wp-block-paragraph">Booking.com&#8217;s experience with Apollo Federation shows how they eliminated API bottlenecks while future-proofing their platform. Through careful design, strong monitoring, and a focus on developer experience, they scaled to billions of daily requests with exceptional performance. </p>



<p class="wp-block-paragraph">To learn more from Sanver Tarmur and Booking.com, watch his full session from GraphQL Summit 2024.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="Making it easier for everyone: How Booking.com securely decomposed their monolith with federation" width="580" height="326" src="https://www.youtube.com/embed/bWC5oM9oa9k?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>



<p class="wp-block-paragraph"></p>


graphic showcasing the title of the blog post

How Booking.com Orchestrated Their Service Architecture with Apollo Federation


<p class="wp-block-paragraph"><kbd>This post is part of our GraphQL Summit 2024 customer series, highlighting how leading teams are using Apollo to accelerate development and build world-class applications. Join hundreds of your peers this October as we return for <a href="https://www.apollographql.com/graphql-summit-2025">GraphQL Summit 2025.</a></kbd></p>



<hr class="wp-block-separator aligncenter has-alpha-channel-opacity" />



<p class="wp-block-paragraph"><a href="https://www.playstation.com/en-us/">PlayStation</a>, a leader in the gaming industry, has transformed its digital store experience by integrating GraphQL into its architecture. Powering experiences across PS5, PlayStation VR, Portal, Mobile, and PlayStation&#8217;s services ecosystem, the platform gives users access to a library of games, including in-house creations like God of War, Spider-Man, and their latest hit, Astrobot. At <a href="https://www.apollographql.com/events/series/graphql-summit-2024">GraphQL Summit 2024</a>, Devashree Shirude, Senior Software Engineer at PlayStation, shared how the company overhauled its digital store experience with federated GraphQL with Apollo, effectively orchestrating multiple backend APIs into a unified, high-performance experience.&nbsp;</p>



<h2 class="wp-block-heading"><a href="#the-api-orchestration-challenge" aria-hidden="true" class="aal_anchor" id="the-api-orchestration-challenge"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The API Orchestration Challenge</h2>



<p class="wp-block-paragraph">When preparing for the PS5 launch, PlayStation examined their tech stack and identified business bottlenecks in its existing REST-based architecture. The system placed heavy computational logic on the client side, leading to duplicate code across platforms and inconsistent user experiences.&nbsp;</p>



<p class="wp-block-paragraph">&#8220;If you&#8217;ve ever bought a game or add-on on PlayStation, you might be wondering what made our stores look so consistent or so robust when you switch between mobile and web platforms,&#8221; Devashree explained.&nbsp;</p>



<p class="wp-block-paragraph">However, achieving this consistency was difficult. Each platform, console, web, and mobile, had to manage its own business logic – a complex web of APIs. Each platform had to independently orchestrate these APIs, resulting in duplicated orchestration logic, redundant implementations, and data inconsistencies across devices.&nbsp;</p>



<p class="wp-block-paragraph">What PlayStation needed was a centralized, declarative approach to API orchestration that could handle the complex business logic while allowing each platform to maintain its unique experience.</p>



<h2 class="wp-block-heading"><a href="#the-path-to-graphql" aria-hidden="true" class="aal_anchor" id="the-path-to-graphql"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The Path to GraphQL</h2>



<p class="wp-block-paragraph">The shift to GraphQL was guided by a core principle, as Devashree noted: &#8220;At PlayStation, we believe that every application should offer a unique experience to our users.&#8221; Instead of forcing a one-size-fits-all approach, PlayStation used GraphQL to create a unified data layer, allowing platforms to maintain customized user experiences while leveraging a shared backend.</p>



<h3 class="wp-block-heading"><a href="#implementation-strategy" aria-hidden="true" class="aal_anchor" id="implementation-strategy"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Implementation Strategy</h3>



<p class="wp-block-paragraph"><strong>Centralized API Orchestration</strong></p>



<p class="wp-block-paragraph">GraphQL federation with Apollo provided PlayStation with a declarative approach to API orchestration. Instead of each client writing procedural code to chain API calls, transform responses, and handle errors, these orchestration patterns were defined once in the GraphQL layer. </p>



<p class="wp-block-paragraph">This orchestration layer automatically handles critical concerns like sequencing API calls, parallelizing requests where possible, transforming data into the exact shape needed by clients, and implementing resilience patterns like retries and fallbacks. It provided a consistent foundation while allowing platforms to maintain unique user journeys.</p>



<p class="wp-block-paragraph">One example is the<em> wishlist </em>feature. &#8220;Though the UI looks compact, thousands of computations are happening in the backend,&#8221; shared Devashree. Instead of separate implementations for each platform, GraphQL resolvers now centrally handle wishlist operations, while platforms retain control over UI design and interactions. </p>



<p class="wp-block-paragraph"><strong>Revolutionizing Business Logic</strong></p>



<p class="wp-block-paragraph">One of the most significant improvements came from centralizing complex business computations. &#8220;What I see on the screen would be different from what you see on the screen,&#8221; Devashree explained. These computations take into account various factors including user subscriptions, past purchases, entitlements, and loyalty tier status.</p>



<p class="wp-block-paragraph">By moving these computations from the client layer to GraphQL federation, PlayStation eliminated the need for duplicate code across platforms.</p>



<p class="wp-block-paragraph"><strong>Optimizing Performance with Apollo Client</strong></p>



<p class="wp-block-paragraph">To accelerate performance and enhance user experience, PlayStation adopted Apollo Client 3.7’s <a href="https://www.apollographql.com/docs/react/data/defer">@defer directive</a> to quickly create a progressive loading experience for users. “Rather than loading everything at once, we break down the UI into chunks, ensuring users see important details first. This not only improves speed but also positively impacts revenue,” Devashree noted.&nbsp;</p>



<p class="wp-block-paragraph">With deferred loading, PlayStation saw faster page interactions, improved conversion rates, and better overall store performance.</p>



<h2 class="wp-block-heading"><a href="#best-practices-and-lessons-learned" aria-hidden="true" class="aal_anchor" id="best-practices-and-lessons-learned"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Best Practices and Lessons Learned</h2>



<p class="wp-block-paragraph">Devashree emphasized the importance of cross-functional collaboration: &#8220;Break those barriers, go out there and talk to your cross-functional groups, like your product team, your user experience team, your design team saying that hey, we already have this available, why not build on top of it?&#8221; By encouraging collaboration, PlayStation ensured developers, designers, and product teams aligned on GraphQL capabilities, and new features leveraged existing GraphQL logic instead of reinventing the wheel.</p>



<p class="wp-block-paragraph">However, Devashree cautioned about potential pitfalls: &#8220;be careful of regressions, each app has its own character, each application has its own tech stack&#8230; if something is happening on Windows, not necessarily it&#8217;s gonna happen on iOS because they are two different OSs.&#8221; By maintaining platform-specific flexibility, PlayStation balanced consistency with customization.</p>



<h2 class="wp-block-heading"><a href="#takeaways" aria-hidden="true" class="aal_anchor" id="takeaways"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Takeaways</h2>



<p class="wp-block-paragraph">With GraphQL federation, PlayStation has built a robust foundation that&nbsp; accelerates performance, boosts&nbsp; developer efficiency, and streamlines updates. Mobile updates, which previously required lengthy app store approvals, can now be shipped faster since core logic resides in GraphQL rather than in platform-specific applications.</p>



<p class="wp-block-paragraph">By cutting development time through code reuse, accelerating store performance with smart loading strategies, and maintaining cross-platform consistency while allowing for customization, PlayStation has created a more efficient, scalable digital store experience.</p>



<figure class="wp-block-pullquote"><blockquote><p><em>&#8220;API orchestration used to be a major pain point for us,&#8221; Devashree reflected. &#8220;With federation, we&#8217;ve transformed it into a competitive advantage. Our teams can focus on building great gaming experiences instead of wrestling with the complexities of connecting multiple backend services.&#8221;</em></p></blockquote></figure>



<p class="wp-block-paragraph"></p>



<p class="wp-block-paragraph">To learn more from Devashree and PlayStation&#8217;s journey, watch her full session from GraphQL Summit 2024.</p>



<figure class="wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio"><div class="wp-block-embed__wrapper">
<iframe loading="lazy" title="Defer then React: The secret to optimizing performance and revenue for PlayStation Store" width="580" height="326" src="https://www.youtube.com/embed/a1A6SD-WmZo?feature=oembed" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</div></figure>


How PlayStation Optimized Performance and Revenue for Their Digital Store with GraphQL Federation


<p class="wp-block-paragraph">Realtor.com is a trusted real estate listing platform that connects buyers and realtors for a seamless experience when renting, buying, or selling properties. At <a href="https://www.youtube.com/watch?v=w07Z572q3kg">GraphQL Summit 2023</a>, Suzy Julius and Cabe Waldrop — SVP of Product and Engineering and Principal Engineer at Realtor.com —&nbsp; discussed the company’s GraphQL journey to simplify its tech stack and scale without compromising performance.&nbsp;</p>



<h2 class="wp-block-heading"><a href="#context-and-challenges" aria-hidden="true" class="aal_anchor" id="context-and-challenges"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Context and challenges</h2>



<p class="wp-block-paragraph">In 2020, Suzy and her team performed a technology audit and found that many of their decisions were based on short-term gains, incomplete migrations, and siloed platforms. In response, the company adopted a technology strategy called “simplify to scale” to target new businesses. The involved teams narrowed in on four key areas of focus: data platform, identity, CI/CD pipeline, and the API layer.&nbsp;</p>



<p class="wp-block-paragraph">In order to tackle these challenges, Suzy and her team evaluated many API solutions but ultimately, they decided GraphQL and <a href="https://www.apollographql.com/docs/federation/">Apollo Federation</a> was the best solution to streamline and unify their APIs. Suzy and Cabe found that only addressing the technological components hindered their progress; they needed to invest just as many resources into storytelling and communication to connect with other teams.</p>



<figure class="wp-block-image"><img decoding="async" src="https://lh7-us.googleusercontent.com/J5Phv-NHwB-L3F7r0th1uSAKywt5rYL_C0dKI8TYLp6FKcCo3yMiCtTLQRpx2ugeptbDu7uisYx1qLDsbJrJ9L2_GS8DEZ7WESQMZHkAgYAcODxSHvvGdn8FZEFt6gkfKLML8SYmQTfK--yUulEfRss" alt=""/><figcaption class="wp-element-caption">Figure: Realtor.com journey to GraphQL adoption</figcaption></figure>



<p class="wp-block-paragraph">Read on to learn about Realtor.com’s most prominent requirements to adopt new technologies and solutions throughout the three phases of its journey: the tech stack, the supergraph, and the API paradigm shifts needed to deliver great UI experiences to clients. They discuss each phase through the lens of their four-step process for GraphQL adoption resiliency, which include solving the technical problem, building momentum, overcoming resistance, and communicating at every level. Let’s dive in.</p>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="522" src="https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-1024x522.png" alt="" class="wp-image-23655" srcset="https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-1024x522.png 1024w, https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-300x153.png 300w, https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-768x391.png 768w, https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-1536x782.png 1536w, https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM-1200x611.png 1200w, https://wp.apollographql.com/wp-content/uploads/2024/04/Screen-Shot-2024-04-22-at-9.48.32-AM.png 1908w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption">Figure: Four-step process to create adoption resiliency </figcaption></figure>



<h2 class="wp-block-heading"><a href="#the-road-to-graphql-adoption-with-apollo" aria-hidden="true" class="aal_anchor" id="the-road-to-graphql-adoption-with-apollo"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The road to GraphQL adoption with Apollo</h2>



<h3 class="wp-block-heading"><a href="#tech-stack" aria-hidden="true" class="aal_anchor" id="tech-stack"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Tech stack&nbsp;</h3>



<p class="wp-block-paragraph">The first area that Realtor.com targeted was the API layer. Realtor.com had previously struggled with over-indexed microservices, endpoint and API inconsistencies, and complex data aggregation on the client side. Cabe and his team considered many options — like service mesh, microservices, and shared libraries — but ultimately, they decided GraphQL was the best solution to streamline and unify their APIs.&nbsp;</p>



<p class="wp-block-paragraph">Suzy and Cabe quickly realized that to implement a unifying API solution, they needed to collaborate with other teams to create understanding and transparency.&nbsp;</p>



<p class="wp-block-paragraph">Suzy and Cabe created a Strategic Technical Initiative Group (STIG) to approach the transition with multiple individuals and teams, including executive sponsors, tech drivers, working groups, and stakeholders. Each of these groups provided different areas of expertise to help the GraphQL STIG prepare for opposition.&nbsp;</p>



<p class="wp-block-paragraph">It came as no surprise that the STIG faced resistance from engineering, product, design, leadership, and cross-functional teams. “Don&#8217;t assume that everyone understands or cares about your technical challenges,” said Cabe. “People aren&#8217;t going to be swayed because you&#8217;ve lined the boxes up nicely… but the good news is that you can get through [to them] if you persevere.”&nbsp;</p>



<p class="wp-block-paragraph">Suzy continued, “We really leaned into the idea that your single biggest problem in communication is the illusion that it&#8217;s already taken place.”&nbsp;</p>



<p class="wp-block-paragraph">With that in mind, the two began tailoring their messaging to each individual team, appealing to their specific interests and challenges. This approach helped the STIG better communicate GraphQL advantages like a single endpoint for all clients, standard payload and error handling, and standardized authentication. As those messages were received, teams started to understand how GraphQL would solve issues they faced day-to-day, related to complexity, security, and developer quality of life.</p>



<figure class="wp-block-image"><img decoding="async" src="https://lh7-us.googleusercontent.com/DqmvjUCOOYAKJVtW2NW_ctHQGTYPxq4s8kNWfL6uJePSWYIzteDt1UzLdxzYjsGOuUpCfzIHTV3blqvQIbiCYVi_6qAB-jnogaqA1m9UPRew_8gVkgg3sLYGlxQCsrnaHekWXtsfm0TX8GlHVPMF5Es" alt=""/><figcaption class="wp-element-caption">Figure: The GraphQL advantage</figcaption></figure>



<h3 class="wp-block-heading"><a href="#the-supergraph" aria-hidden="true" class="aal_anchor" id="the-supergraph"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The supergraph&nbsp;</h3>



<p class="wp-block-paragraph">Once the STIG sold Realtor.com teams on a GraphQL migration, they faced another challenge: rebuilding a monolithic schema. The existing centralized ownership model had lengthy review processes, unclear roles, and inadequate tools that were slowing progress. They found themselves in a position where too many actors were editing and handling the code base, contributing to a fractured API sprawl.&nbsp;</p>



<p class="wp-block-paragraph">The core GraphQL team decided to employ <a href="https://www.apollographql.com/docs/federation/">Apollo Federation</a> to achieve a federated GraphQL architecture called the <a href="https://www.apollographql.com/supergraph/">supergraph</a>. Federation maintains the simplicity of the monolith for client teams but offers a modular, decoupled approach for service teams — allowing for individual GraphQL APIs maintained by separate teams on separate servers called <a href="http://r">subgraphs</a>. These subgraphs sit behind a single API router and access point and combine into one schema to form a supergraph.</p>



<p class="wp-block-paragraph"><strong>Under the supergraph, each service team at Realtor.com could maintain and release changes on their own GraphQL APIs and define core entities in one graph. Plus, client developers could consume data from the federated graph at a single endpoint, quickly and painlessly.</strong></p>



<p class="wp-block-paragraph">To support this process, Suzy and Cabe recommend finding early adopters and providing them with resources to get the supergraph project off the ground. At Realtor.com, security and growth teams were particularly excited about GraphQL, so the core team built a guild to discuss best practices. When they eventually faced opposition about moving away from a monolith approach, they had already focused on building out the tools and automation that simplify subgraph deployment for better visualization.</p>



<p class="wp-block-paragraph">Suzy said, “Big migrations are often a combination of focusing on the journey and not just the destination, but also celebrating the wins when you have them. Let&#8217;s be honest, it takes a bit of cleverness to get things done right.”</p>



<h3 class="wp-block-heading"><a href="#paradigm-shifts" aria-hidden="true" class="aal_anchor" id="paradigm-shifts"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Paradigm shifts</h3>



<p class="wp-block-paragraph">The major paradigm shift that Realtor.com faced involved finding a way for its APIs to both express the intent of the product and power a high-quality user experience. Realtor.com displays listings for thousands of homes, with specific information about features, timelines, and price drops. Developer teams need to write the same code for every single one of these experiences across Android, iOS, web, and more — and as the number of clients increases, the likelihood that the code stays uniform decreases.&nbsp;</p>



<p class="wp-block-paragraph">“This is where GraphQL really shines, regardless of whether you&#8217;re in tech, product, or design,” Cabe said.&nbsp;</p>



<p class="wp-block-paragraph">GraphQL allows teams to get features out the door faster and ensure quality across an entire product suite by building experience-driven APIs and moving logic to the backend. Furthermore, with a single backend expression of core business logic, it streamlines how teams understand and process data. These benefits support a focus on the entire product and a stronger sense of customer trust.</p>



<p class="wp-block-paragraph">“This is more of a cultural shift than a technical problem you&#8217;re solving for,” Suzy clarified. “We can&#8217;t stress enough just how critical it is that you need to build resiliency into your business because if you do that, and you communicate, you can continue this big paradigm shift when business is up and also when it&#8217;s down.”&nbsp;</p>



<p class="wp-block-paragraph">Suzy and Cabe said this comes down to exposure. Paradigm shifts are about noting what works when showing different teams how GraphQL impacts business processes and results.</p>



<h2 class="wp-block-heading"><a href="#looking-toward-the-future" aria-hidden="true" class="aal_anchor" id="looking-toward-the-future"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Looking toward the future</h2>



<p class="wp-block-paragraph">By jumping headfirst into GraphQL with Apollo, Realtor.com transitioned from multiple product expressions to a shared product expression, from an architectural point of technical leverage to an organizational point of product leverage, and from many services and data sources to a single platform.&nbsp;</p>



<p class="wp-block-paragraph">Cabe emphasized, “If you&#8217;re an engineer, there&#8217;s a level of maturity and growth that comes from investing in the product side and knowing more about what you&#8217;re trying to deliver. Think about those APIs and about how they power the experience so you can deliver more in the future.”&nbsp;</p>



<p class="wp-block-paragraph">Check out Suzy Julius and Cabe Waldrop’s insightful <a href="https://www.youtube.com/watch?v=w07Z572q3kg&amp;list=PLpi1lPB6opQzUOqG3QroLLN06FF-Q_uhX&amp;index=3">session at the GraphQL Summit 2023</a> to learn more. Ready to start your GraphQL journey with Apollo? <a href="https://www.apollographql.com/contact-sales?referrer=nav-contact-sales">Contact</a> one of our representatives today!&nbsp;</p>


How Realtor.com Achieved Technological and Cultural Transformation with GraphQL


<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="573" src="https://wp.apollographql.com/wp-content/uploads/2024/04/API-Days-NY-Apollo-GraphQL-Speakers-1024x573.png" alt="" class="wp-image-23618" srcset="https://wp.apollographql.com/wp-content/uploads/2024/04/API-Days-NY-Apollo-GraphQL-Speakers-1024x573.png 1024w, https://wp.apollographql.com/wp-content/uploads/2024/04/API-Days-NY-Apollo-GraphQL-Speakers-300x168.png 300w, https://wp.apollographql.com/wp-content/uploads/2024/04/API-Days-NY-Apollo-GraphQL-Speakers-768x430.png 768w, https://wp.apollographql.com/wp-content/uploads/2024/04/API-Days-NY-Apollo-GraphQL-Speakers.png 1186w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<p class="wp-block-paragraph">API Days NY is a must-attend for those looking to network with API leaders and learn how top enterprises are leveraging federated GraphQL in their API platforms. As you can see from the agenda highlights below, it will be an exciting two day event.&nbsp;</p>



<ul class="wp-block-list">
<li>Our founder and CTO, Matt DeBergalis, will share his keynote &#8220;The Year Ahead for GraphQL Federation: Uniting APIs for the AI Era”</li>



<li>API leaders from Netflix, Wayfair and Apollo will present sessions on optimizing performance, creating effective api schemas and building an internal community for effective API governance</li>



<li>Apollo is offering a free workshop to guide you through building your first GraphQL query and beyond</li>



<li><a href="https://bit.ly/43IhuX4">Register to join us for a meetup &amp; happy hour</a> to meet with peer API leaders and GraphQL experts</li>
</ul>



<p class="wp-block-paragraph">Swing by the Apollo booth to connect, and read on for more details.</p>



<h3 class="wp-block-heading"><a href="#keynote-the-year-ahead-for-graphql-federation-combining-apis-for-the-ai-era" aria-hidden="true" class="aal_anchor" id="keynote-the-year-ahead-for-graphql-federation-combining-apis-for-the-ai-era"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Keynote: The Year Ahead for GraphQL Federation: Combining APIs for the AI Era</strong></h3>



<p class="wp-block-paragraph"><em>Matt DeBergalis, Founder and CTO, Apollo GraphQL</em></p>



<p class="wp-block-paragraph">Forget what you’ve heard about GraphQL – it’s not a REST alternative, graph database, or just for front-end developers. GraphQL is about combining the APIs you already have.</p>



<p class="wp-block-paragraph">In this talk, you’ll learn how a federated GraphQL architecture brings a platform approach to API composition, where it fits into the API stack, and what is coming next to help teams build tomorrow’s AI products and experiences on today’s APIs.</p>



<h3 class="wp-block-heading"><a href="#session-the-graphql-overhead-myth-or-reality" aria-hidden="true" class="aal_anchor" id="session-the-graphql-overhead-myth-or-reality"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Session: The GraphQL Overhead: Myth or Reality?</strong></h3>



<p class="wp-block-paragraph"><em>Marc-André Grioux, Senior Software Engineer &amp; Author, Netflix</em></p>



<p class="wp-block-paragraph">GraphQL is a powerful tool to enable declarative data fetching on the client and dynamic data representation on the server. This power must have an overhead. In this talk, investigate what this overhead is made of, how impactful it can be to scaling an API platform and services, and how to mitigate it as much as possible while retaining the benefits of GraphQL.</p>



<h3 class="wp-block-heading"><a href="#session-the-secrets-to-graph-success-from-inception-to-general-availability" aria-hidden="true" class="aal_anchor" id="session-the-secrets-to-graph-success-from-inception-to-general-availability"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Session: The Secrets to Graph Success, from Inception to General Availability</strong></h3>



<p class="wp-block-paragraph"><em>Leah Hurwich-Adler, Senior Product Leader</em></p>



<p class="wp-block-paragraph">Building an API platform for a Fortune 100 retailer, capable of handling over 600 million requests per day, requires a highly distributed ecosystem and Federated GraphQL. Drawing from firsthand experience, Leah will share insights on building an internal API community, setting up governance models, and igniting leaders&#8217; enthusiasm for your platform vision. Join this talk to uncover the keys to launching a successful GraphQL platform, through lessons that merge technology with strategic and leadership insights.</p>



<h3 class="wp-block-heading"><a href="#session-benefits-of-idiomatic-code-first-graphql-development" aria-hidden="true" class="aal_anchor" id="session-benefits-of-idiomatic-code-first-graphql-development"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Session: Benefits of Idiomatic Code First GraphQL Development</strong></h3>



<p class="wp-block-paragraph"><em>Derek Kuc, Principal Engineer, Apollo GraphQL</em></p>



<p class="wp-block-paragraph">When building a GraphQL server, two main approaches exist: starting with an SDL and matching resolver logic, or generating the schema directly from code. In this talk, explore the benefits of idiomatic code-first GraphQL development, focusing on GraphQL Kotlin examples. Discover how this approach enhances developer experience and API consistency.</p>



<h3 class="wp-block-heading"><a href="#workshop-build-an-api-composition-layer-with-apollo-graphql" aria-hidden="true" class="aal_anchor" id="workshop-build-an-api-composition-layer-with-apollo-graphql"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Workshop: Build an API Composition Layer with Apollo GraphQL</strong></h3>



<p class="wp-block-paragraph"><em>Michael Berger, Solutions Engineer, Apollo GraphQL</em></p>



<p class="wp-block-paragraph">Use GraphQL to accelerate developer velocity by creating a composition layer for your microservices. Orchestrate data by creating a query that spans your REST APIs and delivers exactly the data your application requires. Learn how to build your first GraphQL query and why GraphQL is valuable to your microservice architecture.</p>



<h3 class="wp-block-heading"><a href="#nyc-meetup-and-happy-hour" aria-hidden="true" class="aal_anchor" id="nyc-meetup-and-happy-hour"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>NYC Meetup and Happy Hour</strong></h3>



<p class="wp-block-paragraph"><a href="https://bit.ly/43IhuX4">Register to join the Apollo team, Graph Champions, and apidays attendees</a> on April 30 at 6 PM following the conference for bites and beverages at nearby restaurant The Reading Room (16 E 46th St, New York, NY 10017 &#8211; located in the Gotham Hotel)</p>



<h3 class="wp-block-heading"><a href="#still-deciding-get-your-free-pass-now-before-they-run-out" aria-hidden="true" class="aal_anchor" id="still-deciding-get-your-free-pass-now-before-they-run-out"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>Still deciding? Get your free pass now before they run out!&nbsp;&nbsp;</strong></h3>



<p class="wp-block-paragraph">The full conference ticket includes access to the Keynotes, Track Sessions, Workshops and Networking Breakfasts and Lunches on April 30 &amp; May 1. Email <a href="mailto:champions@apollographql.com">champions@apollographql.com</a> to learn more and request a pass.</p>


Join Us at API Days in New York City, April 30 & May 1, 2024

How Wayfair achieved GraphQL success with federated GraphQL


<h2 class="wp-block-heading"><a href="#industry-leaders-to-showcase-tangible-business-impact-of-graphql-federation" aria-hidden="true" class="aal_anchor" id="industry-leaders-to-showcase-tangible-business-impact-of-graphql-federation"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><em>Industry Leaders to Showcase Tangible Business Impact of GraphQL Federation</em></h2>



<p class="wp-block-paragraph">In the wake of a recent report by Gartner<sup>®</sup>, <a href="https://apollographql.com">Apollo GraphQL</a>, the creator of GraphQL federation, is announcing the 7th annual GraphQL Summit, scheduled for October 8–10, 2024, in New York City. This premier three-day event gathers the largest assembly of GraphQL enthusiasts and API platform teams. Engineering teams can learn from industry leaders spearheading successful GraphQL deployments at scale and hear from Apollo GraphQL on its product roadmap and vision for the future of GraphQL and GraphQL federation.</p>



<p class="wp-block-paragraph">According to the Gartner report, GraphQL&#8217;s adoption is on the rise, since “By 2027, more than 60% of enterprises will use GraphQL in production, up from less than 30% in 2024.” The report also predicts that “By 2027, 30% of enterprises using GraphQL will utilize GraphQL federation, up from less than 5% in 2024.”1<sup data-fn="5626a7a6-afc4-465e-84b4-2219decdae0e" class="fn"><a href="#5626a7a6-afc4-465e-84b4-2219decdae0e" id="5626a7a6-afc4-465e-84b4-2219decdae0e-link">1</a></sup> As GraphQL federation marks its<a href="https://www.apollographql.com/blog/celebrating-five-years-of-apollo-federation"> 5th anniversary,</a> Apollo GraphQL remains at the forefront of this transformative technology.</p>



<p class="wp-block-paragraph">Previous GraphQL Summits have featured internet giants like Netflix, Intuit, and Wayfair, sharing invaluable insights into deploying GraphQL at scale to support distributed architectures. Rather than a sprawl of backends-for-frontends or experience APIs, these teams implemented GraphQL federation, an architecture that combines APIs and enables any number of client teams to query them from a single endpoint. Engineering leaders and API platform teams can learn how to use this architecture to drive developer velocity, enhance the customer experience across any number of applications, or handle the technical side of mergers and acquisitions, where IT is saddled with many disparate siloed systems that were never designed to talk to each other.</p>



<p class="wp-block-paragraph">This year’s GraphQL Summit aims to support this burgeoning adoption with sessions led by industry luminaries across various sectors, including retail, automotive, and banking.&nbsp; Featured speakers include Stephanie Saunders, Engineering Manager at Coinbase; Mike Cohen, Technical Fellow at Indeed; and Marc-Andre Giroux, Senior Software Developer at Netflix, as well as the co-founders of Apollo GraphQL, Geoff Schmidt and Matt DeBergalis, who will share their insights and expertise on the future of GraphQL and GraphQL federation. Attendees can expect tracks and workshops covering best practices on security, governance, and API platform engineering; real-world use cases and the latest advancements in GraphQL technology; along with a new focus on leveraging GraphQL federation to support artificial intelligence.&nbsp;</p>



<figure class="wp-block-pullquote"><blockquote><p>Geoff Schmidt, CEO and co-founder of Apollo GraphQL, remarks: “As enterprise technology evolves, the rise of GraphQL represents a pivotal shift towards more efficient, scalable, and interconnected systems. The latest insights from Gartner reaffirm what we at Apollo GraphQL have long believed: GraphQL federation is not just a trend, but a fundamental technology shaping the future of digital experiences. Apollo GraphQL is committed to fostering an environment where developers and enterprises can learn, grow, and succeed with GraphQL. The upcoming GraphQL Summit is a reflection of this commitment, featuring talks from industry leaders like Coinbase and Netflix, and promises to be an invaluable opportunity for API platform engineers and leaders to discover how GraphQL federation revolutionizes API architectures for seamless, innovative systems that are driving the future of digital experiences.”</p></blockquote></figure>



<p class="wp-block-paragraph">All GraphQL enthusiasts are invited to join the conversation by submitting to speak at the Summit. Whether an industry veteran or a newcomer, this is a chance to share insights, experiences, and innovations with the GraphQL community.</p>



<p class="wp-block-paragraph">For more information about the GraphQL Summit or to submit your proposal to speak, please visit our website: <a href="http://www.graphqlsummit.com">www.graphqlsummit.com</a>.</p>



<h2 class="wp-block-heading"><a href="#resources" aria-hidden="true" class="aal_anchor" id="resources"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Resources</h2>



<ul class="wp-block-list">
<li>Read the<a href="https://www.apollographql.com/resources/gartner-when-to-use-graphql-to-accelerate-api-delivery"> latest Gartner report</a>, When to Use GraphQL to Accelerate API Delivery</li>



<li>Learn more about <a href="https://www.apollographql.com/blog/celebrating-five-years-of-apollo-federation">Apollo Federation</a>, Apollo’s implementation of GraphQL Federation&nbsp;</li>



<li>Stay up to date on the <a href="http://www.graphqlsummit.com">GraphQL Summit </a>agenda and learn more about it <a href="https://www.apollographql.com/blog/announcing-graphql-summit-2024-in-new-york-city">here</a></li>



<li><a href="https://www.apollographql.com/graphos/">Build </a>a GraphQL API platform today</li>



<li><a href="https://www.apollographql.com/tutorials/certifications/apollo-graph-professional/learning-path/">Become certified </a>as an Apollo GraphQL developer</li>



<li><a href="https://discord.com/invite/yFZJH2QYrK">Join</a> the Apollo GraphQL Community on Discord</li>
</ul>



<h2 class="wp-block-heading"><a href="#about-apollo-graphql" aria-hidden="true" class="aal_anchor" id="about-apollo-graphql"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>About Apollo GraphQL</h2>



<p class="wp-block-paragraph"><a href="https://apollographql.com">Apollo GraphQL </a>helps developers help the world by providing a better API platform across teams. Apollo’s open-source software is downloaded 25M times per month and its commercial GraphQL technologies power the most innovative brands today. Teams at Netflix, Snap, and the New York Times ship personalized, omnichannel experiences faster with a supergraph – a self-service GraphQL platform that spans any number of backend services. Serving over 5T requests in 2023, the Apollo GraphOS<sup>TM</sup> platform simplifies API development with workflows and infrastructure to build, test, and ship supergraphs at any scale. Based in San Francisco, Apollo is backed by Insight Partners, Andreessen Horowitz, Matrix Partners, and Trinity Ventures. Learn more at: <a href="https://www.apollographql.com">https://www.apollographql.com</a>.</p>


<ol class="wp-block-footnotes"><li id="5626a7a6-afc4-465e-84b4-2219decdae0e">Gartner, When to Use GraphQL to Accelerate API Delivery, By Shameen Pillai, et al, March 7, 2024. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved. <a href="#5626a7a6-afc4-465e-84b4-2219decdae0e-link" aria-label="Jump to footnote reference 1">↩︎</a></li></ol>

Image portays GraphQL Summit by Apollo logo and includes text that reads the conference for API platform engineering builders and leaders. October 8-10, 2024. Convene 225 Liberty New York City.

Apollo GraphQL Announces the Next GraphQL Summit Amid Rapid GraphQL Enterprise Adoption 


<p class="wp-block-paragraph">The Netflix API evolution is a saga worth exploring, spanning 15 years and five generations of APIs, culminating in GraphQL adoption to meet the growing demands of its industry-leading streaming platform. Bruce Wang, Director of Product Platform Systems at Netflix, sat down with Matt DeBergalis, CTO at Apollo GraphQL, to discuss Netflix&#8217;s extensive API journey and the adoption of <a href="https://www.apollographql.com/docs/resources/faq/#what-is-graphql">GraphQL</a> in an insightful fireside discussion during the <a href="https://www.youtube.com/watch?v=b78Q7qY4IpI&amp;t=121s">GraphQL Summit 2023</a>.</p>



<figure class="wp-block-embed is-type-video is-provider-wistia-inc wp-block-embed-wistia-inc"><div class="wp-block-embed__wrapper">
<script src="https://fast.wistia.com/assets/external/E-v1.js" async></script><div class="wistia_responsive_padding" style="padding:56.25% 0 0 0;position:relative;"><div class="wistia_responsive_wrapper" style="height:100%;left:0;position:absolute;top:0;width:100%;"><div class="wistia_embed wistia_async_gx4gps0gg9 dnt=1 seo=false videoFoam=true" style="height:100%;position:relative;width:100%"><div class="wistia_swatch" style="height:100%;left:0;opacity:0;overflow:hidden;position:absolute;top:0;transition:opacity 200ms;width:100%;"><img decoding="async" src="https://fast.wistia.com/embed/medias/gx4gps0gg9/swatch" style="filter:blur(5px);height:100%;object-fit:contain;width:100%;" alt="" aria-hidden="true" onload="this.parentNode.style.opacity=1;" /></div></div></div></div>
</div></figure>



<h2 class="wp-block-heading"><a href="#netflixs-api-evolution-the-unexpected-journey" aria-hidden="true" class="aal_anchor" id="netflixs-api-evolution-the-unexpected-journey"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Netflix&#8217;s API Evolution: The Unexpected Journey</h2>



<p class="wp-block-paragraph">Their conversation built on Bruce’s earlier webinar, &#8220;<a href="https://www.apollographql.com/events/champions-corner/thank-you/netflixs-unexpected-journey-to-the-supergraph/">Netflix APIs: An Unexpected Journey</a>&#8221; where he shared how their multi-year, multi-generation API evolution culminated in a new &#8220;Consumer Edge&#8221; powered by federated GraphQL.<br><br>Along the way the Netflix API team adopted and then outgrew multiple API architectures. From their initial catalog of distributed REST endpoints coined “OpenAPI&#8221; to its replacement API platform “API.next” and then to “DNA” a GraphQL-like API. In each case, they were encased in a complex monolith, so their goal was threefold: break the monolithic architecture, unify the APIs across diverse platforms and empower domain owners to manage their APIs independently.&nbsp; </p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph">“In all our previous stacks, each individual UI built their own APIs. iOS had one API, Android had another API, TV had a different API, etc. we just couldn’t scale with the business anymore. Unifying those in a single federated GraphQL API was super valuable, and something we couldn&#8217;t do before.”</p>



<p class="wp-block-paragraph">&#8211; <em>Bruce Wang, Director of Product Platform Systems team, Netflix</em></p>
</blockquote>



<p class="wp-block-paragraph">As described in their 2020 blog <a href="https://netflixtechblog.com/how-netflix-scales-its-api-with-graphql-federation-part-1-ae3557c187e2">How Netflix Scales its API with GraphQL Federation</a>, the Netflix API team found <a href="https://www.apollographql.com/docs/federation/subgraph-spec/">the Apollo Federation specification</a> the perfect way to scale their GraphQL architecture. In the federated model, individual GraphQL schemas become subgraphs, which are composed into a unified supergraph. In this way, they retained the integrated “Consumer Edge” API they sought while decoupling their many domain teams. The result was faster delivery without sacrificing a cohesive customer experience.&nbsp;</p>



<p class="wp-block-paragraph">The graph&#8217;s power to democratize information also helped foster data-centric discussions, empowering every Netflix team member to meaningfully engage using concrete information.&nbsp; </p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>&#8220;There have been numerous instances where people ask, &#8216;What does this subgraph do?&#8217; And now we can simply go in and see it all, what it does, who calls it. That sounds basic, but we lacked this information with previous APIs; it was like dealing with a black box. Only the API team knew how to manage that black box.&#8221;</em></p>



<p class="wp-block-paragraph">&#8211; <em>Bruce Wang, Director of Product Platform Systems team, Netflix</em></p>
</blockquote>



<p class="wp-block-paragraph">For a full recounting of Netflix&#8217;s API transformation journey, be sure to watch Bruce’s &#8220;<a href="https://www.apollographql.com/events/champions-corner/thank-you/netflixs-unexpected-journey-to-the-supergraph/">Netflix APIs: An Unexpected Journey</a>&#8221; webinar.&nbsp; </p>



<h2 class="wp-block-heading"><a href="#strategic-lessons-from-netflixs-api-evolution" aria-hidden="true" class="aal_anchor" id="strategic-lessons-from-netflixs-api-evolution"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Strategic Lessons from Netflix&#8217;s API Evolution</h2>



<p class="wp-block-paragraph">Bruce expanded on Netflix&#8217;s API journey at a fireside chat with Matt at GraphQL Summit 2023. In this talk, Bruce highlighted some helpful tips for rolling out a graph across a broader organization, emphasizing the strategic approach of tackling complex challenges first, balancing the needs of API clients and service owners, leaning into developer experience and API governance, and discussing the human element for successful technological innovation. Read on to learn more.&nbsp;</p>



<h3 class="wp-block-heading"><a href="#lesson-1-tackle-complex-challenges-first" aria-hidden="true" class="aal_anchor" id="lesson-1-tackle-complex-challenges-first"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Lesson 1: Tackle Complex Challenges First</h3>



<p class="wp-block-paragraph">When undertaking major architectural transformations, a conventional approach often suggests starting simple and working up to complex use cases. However, when Matt asked about the key lessons from the API transformative journey, Bruce highlighted the strategy of addressing the most complex customer experiences first, which for Netflix would have been its homepage. This approach helps in de-risking the project and provides immediate value, testing the robustness of the new API strategy under demanding conditions.&nbsp;</p>



<p class="wp-block-paragraph">In discussing this, Bruce emphasized that technology choices must deliver tangible user benefits, emphasizing the essence of Netflix&#8217;s strategic technological shifts anchored in delivering meaningful outcomes.&nbsp; </p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>&#8220;For me, I really believe in providing value. I don&#8217;t believe in adopting technology just because it&#8217;s cool. It&#8217;s cool, but it has to provide value, right?”</em></p>



<p class="wp-block-paragraph">&#8211; <em>Bruce Wang, Director of Product Platform Systems team, Netflix</em></p>
</blockquote>



<h3 class="wp-block-heading"><a href="#lesson-2-balance-the-needs-of-api-clients-and-service-owners" aria-hidden="true" class="aal_anchor" id="lesson-2-balance-the-needs-of-api-clients-and-service-owners"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Lesson 2: Balance the Needs of API Clients and Service Owners&nbsp;</h3>



<p class="wp-block-paragraph">While GraphQL is often associated with frontend development due to its client-centric approach, its true value extends far beyond the frontend.&nbsp;</p>



<p class="wp-block-paragraph">During the conversation, Bruce highlighted that although his team initially centered on the API client&#8217;s value proposition, the focus shifted towards federating and democratizing GraphQL development&nbsp; over time. This evolution stemmed from the necessity to break down a complex monolith across major domains such as the growth graph, discovery graph, and identity graph. The API team aimed to distribute ownership, allowing service teams to manage their APIs independently.&nbsp;</p>



<p class="wp-block-paragraph">An unintended consequence of Netflix&#8217;s journey into microservices was the API team inadvertently becoming the central clearing house for an intricate web of models and dependencies, crucial for diverse functionalities. The organic development of graphs within backend and frontend teams marked a significant shift, reducing dependency on the API team. When Matt inquired about the value of this approach, Bruce emphasized that it fostered independent evolution, creating an environment where new graph components emerged organically without the API team&#8217;s direct involvement.&nbsp;</p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“ [&#8230;] the thing that we really wanted is for the UI team to talk to the backend team without us being in the middle. [&#8230;] The design of the graphs are happening organically; the back-end teams and the UI teams are working together. We&#8217;re using tooling to make the sharing easier, the testing easier. New graphs are coming up without us even knowing.”&nbsp;</em></p>



<p class="wp-block-paragraph">&#8211; <em>Bruce Wang, Director of Product Platform Systems team, Netflix</em></p>
</blockquote>



<p class="wp-block-paragraph">While this heralds a new era of dynamic API development, Bruce cautioned that the challenge remains in balancing autonomy without federating complexity at Netflix. He noted that the API team held significant expertise in managing a large-scale graph, and transferring this responsibility abruptly to service teams could create complications. Bruce emphasized that maintaining the consolidated experience of the API team is crucial to prevent potential issues arising from distributing this responsibility among multiple backend teams.&nbsp;</p>



<h3 class="wp-block-heading"><a href="#lesson-3-lean-into-developer-experience-api-governance" aria-hidden="true" class="aal_anchor" id="lesson-3-lean-into-developer-experience-api-governance"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Lesson 3: Lean into Developer Experience &amp; API Governance</h3>



<p class="wp-block-paragraph">The discussion also touched on governance aspects in API management. Bruce highlighted a specific instance where the challenge arose while migrating intricate iOS and Android-specific APIs into a more universally applicable graph. This process revealed significant logic embedded in the UI code and Backend for Frontend (BFFs), necessitating strategic decisions on representation and ownership within the graph. The API team intentionally took the lead in the initial translation to simplify these complexities, paving the way for future domain ownership. Bruce elaborates on their collaborative approach with sister teams, such as the Consumer Identity and Access (CIA) team, indicating an ongoing phase where the API team currently manages while actively facilitating a transition of ownership.</p>



<p class="wp-block-paragraph">The Netflix journey portrays deliberate steps in crafting a schema framework, navigating complexities in transitioning logic to the graph, and fostering a decentralized governance model for effective ownership and evolution of APIs.</p>



<h3 class="wp-block-heading"><a href="#lesson-4-bridge-silos-by-fostering-inter-team-conversations" aria-hidden="true" class="aal_anchor" id="lesson-4-bridge-silos-by-fostering-inter-team-conversations"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>Lesson 4: Bridge Silos by Fostering Inter-Team Conversations</h3>



<p class="wp-block-paragraph">In the realm of Netflix&#8217;s API evolution, the conversation steered into the intricacies of the <a href="https://www.apollographql.com/docs/federation/#next-steps">supergraph</a> and its profound impact on organizational dynamics and developer experiences. Bruce explained that while they employ advanced tooling to facilitate graph management, they also invest significantly in fostering intentional communication patterns among teams.</p>



<p class="wp-block-paragraph">This emphasis on meaningful dialogue arises from Bruce’s acknowledgment that within organizational frameworks, teams can easily retreat into isolated spheres. Bruce recognized this issue, pointing out that many teams might opt to function within their silos and avoid conversations between teams. To overcome this, the company actively encourages these discussions, leveraging the graph as a focal point. Teams are prompted to deliberate on essential matters like whether a growth API should access a discovery API or develop their own, fostering conversations that might otherwise not take place.&nbsp;</p>



<p class="wp-block-paragraph">In essence, this dialogue underscores Netflix&#8217;s ongoing pursuit to blend technological innovation with human interaction—leveraging tools while acknowledging the importance of human communication in steering effective developer experiences.</p>



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
<p class="wp-block-paragraph"><em>“[..] the developer experience to me is a multi-pronged approach [&#8230;] sometimes you&#8217;re using [&#8230;] tooling but sometimes you&#8217;re using just good old fashioned people and trying to nudge the conversation to happen.”</em></p>



<p class="wp-block-paragraph">&#8211; <em>Bruce Wang, Director of Product Platform Systems team, Netflix</em></p>
</blockquote>



<h2 class="wp-block-heading"><a href="#the-journey-ahead" aria-hidden="true" class="aal_anchor" id="the-journey-ahead"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>The Journey Ahead</h2>



<p class="wp-block-paragraph">In the concluding segment of the fireside chat, Matt and Bruce explored the future prospects and challenges for the Netflix API evolution. Bruce reflected on his personal journey, having transitioned from leading the API team to tackling broader, more complex projects at Netflix, such as managing tech debt and orchestrating multiple massive migrations beyond GraphQL. While the transition to GraphQL stands as a pivotal point, Bruce&#8217;s attention now gravitates toward rearchitecting the client telemetry framework, underscoring the commitment to refining and enhancing Netflix&#8217;s technical landscape. Embracing these lessons from the API journey, Bruce emphasized the importance of applying these insights across the organization, aiming to support other teams while navigating their respective challenges.&nbsp;</p>



<p class="wp-block-paragraph">While sharing final thoughts and reflections, Matt commended the environment fostered at Netflix, describing it as an ideal crucible for technological ideas to thrive. He highlighted the unique blend of culture, product excellence, and scale that has propelled the company forward. He hoped sharing Netflix&#8217;s approach would be valuable for others on a similar path.&nbsp;</p>



<p class="wp-block-paragraph">Bruce acknowledged that while Netflix has a strong brand, it still faces challenges and constraints like any company. Despite its size, Netflix doesn&#8217;t have infinite resources. Bruce wanted to &#8220;pierce the aura&#8221; around Netflix and convey that they grapple with similar people and design problems that others face when adopting new paradigms.&nbsp;</p>



<p class="wp-block-paragraph">In closing, Matt noted that many of Netflix&#8217;s innovations arose from working within their constraints, such as not being able to rapidly grow their team as the number of services expanded. This forced them to find scalable, efficient solutions. The key takeaway is that while challenging, it is possible to achieve significant efficiency gains through thoughtful cultural, organizational and technical approaches &#8211; even with limited resources.</p>



<p class="wp-block-paragraph">Watch the <a href="https://www.youtube.com/watch?v=b78Q7qY4IpI&amp;t=121s">GraphQL Summit 2023 panel discussion with Netflix</a> to learn more about the remarkable narrative of growth and adaptability within Netflix&#8217;s evolving API landscape.&nbsp;</p>



<p class="wp-block-paragraph"><em>Note: Dan Boerner also&nbsp;contributed to this blog post.&nbsp;</em></p>


Redefining API Strategy: Why Netflix Platform Engineering Chose Federated GraphQL


<p class="wp-block-paragraph">We are thrilled to announce our diversity scholarship for GraphQL Summit 2022! </p>



<p class="wp-block-paragraph">At Apollo, we are focused on building a team that is inclusive of all demographics. We want Summit to be a reflection of that and bring together all perspectives, so we can develop and learn how the supergraph needs to evolve to serve the different needs of our customers and their end customers.</p>



<p class="wp-block-paragraph">A diversity ticket is a fully sponsored (read: free) ticket for an individual in an underrepresented tech group. Underrepresented groups include but are not limited to: women, people of color, LGBTQ+ people, disabled people, and individuals facing economic or social hardships.</p>



<p class="wp-block-paragraph">Do you qualify? If so, fill out <a href="https://forms.gle/FpFRH8r5BJqdcSpo8">this form</a> by Monday, August 29th. The 30 scholarship awardees will be informed on September 2nd.</p>



<p class="wp-block-paragraph">Please note that the diversity ticket does not include the cost of travel or hotel accommodations.</p>



<p class="wp-block-paragraph"><strong>We’re also looking for a diversity sponsor!</strong><br>The scholarships are made possible in a joint effort with our sponsors. If you work at a company (or know one) that supports diversity in tech and would be <strong>interested in sponsoring the scholarship</strong>, simply drop us a line at <a href="mailto:summit@apollographql.com">summit@apollographql.com</a>.</p>


Announcing the 2022 GraphQL Summit diversity tickets


<p class="wp-block-paragraph">Calling all GraphQL developers: GraphQL Summit, the world’s first and largest conference celebrating the GraphQL community is back, and we can’t wait to host you in person in sunny San Diego!</p>



<h3 class="wp-block-heading"><a href="#graphql-summit-will-take-place-at-the-sheraton-san-diego-hotel-marina-october-3-5-2022" aria-hidden="true" class="aal_anchor" id="graphql-summit-will-take-place-at-the-sheraton-san-diego-hotel-marina-october-3-5-2022"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a><strong>GraphQL Summit will take place at the Sheraton San Diego Hotel &amp; Marina October 3-5, 2022</strong></h3>



<p class="wp-block-paragraph">We are aware that Summit falls on several important religious and holy days, including Yom Kippur and Dussehra. When we made the decision to host Summit in person, there were limited dates and availability at conference venues able to host a group our size. Ultimately, it came down to whether or not to host GraphQL Summit during this time or go 100% virtual again. We recognize that some GraphQL community members observe these important days and may choose not to participate in Summit. Talk recordings will be available after the event.</p>



<h3 class="wp-block-heading"><a href="#graphql-summit-call-for-proposals-due-june-29" aria-hidden="true" class="aal_anchor" id="graphql-summit-call-for-proposals-due-june-29"><svg aria-hidden="true" class="aal_svg" height="16" version="1.1" viewBox="0 0 16 16" width="16"><path fill-rule="evenodd" d="M4 9h1v1H4c-1.5 0-3-1.69-3-3.5S2.55 3 4 3h4c1.45 0 3 1.69 3 3.5 0 1.41-.91 2.72-2 3.25V8.59c.58-.45 1-1.27 1-2.09C10 5.22 8.98 4 8 4H4c-.98 0-2 1.22-2 2.5S3 9 4 9zm9-3h-1v1h1c1 0 2 1.22 2 2.5S13.98 12 13 12H9c-.98 0-2-1.22-2-2.5 0-.83.42-1.64 1-2.09V6.25c-1.09.53-2 1.84-2 3.25C6 11.31 7.55 13 9 13h4c1.45 0 3-1.69 3-3.5S14.5 6 13 6z"></path></svg></a>GraphQL Summit: Call for Proposals (due June 29)</h3>



<p class="wp-block-paragraph">Interested in speaking at GraphQL Summit? All talks will be in person this year, and the deadline for talk submissions is June 29. Speakers will be notified of their acceptance by July 8.</p>



<p class="wp-block-paragraph"><a rel="noreferrer noopener" href="https://docs.google.com/forms/d/e/1FAIpQLSefxh14DJbhIIy1meXw6JAiXw_Nv6NTsGBfjmigZCqBs7_FwQ/viewform" target="_blank">Submit your talk proposal &gt;</a></p>



<p class="wp-block-paragraph">We love talks that focus on real developer use cases and teach the audience something new. To help spark some ideas, we&#8217;ve included some suggested talk topics, but please don&#8217;t limit yourself to our list:</p>



<ul class="wp-block-list"><li><strong>Supergraph</strong>: Federated schema design, building a subgraph in Java/Kotlin, Ruby, Go, Python, Rust, and C#, customizing the Apollo Router, caching, onboarding new teams, sharing types between subgraphs, breaking up a monolith, understanding query plans.</li><li><strong>Best practices</strong>: Schema design, demand control, testing, caching, TypeScript codegen, defer/stream, schema change management, authentication and authorization, error handling, CI/CD, observability, data governance, serverless deployments.</li><li><strong>Client</strong>: SSR, modern React (Suspense and Server Components), client performance, client caching and normalization, pagination, mocking, local state management, networking, iOS and Android clients.</li><li><strong>Future of GraphQL</strong>: New proposals to the GraphQL spec, new GraphQL compilers, interesting use cases (Web3, iOT, AR/VR, data science).</li></ul>



<p class="wp-block-paragraph"><strong>Benefits: </strong>All speakers will receive a conference ticket and 1 buddy pass, an invitation to an exclusive speakers&#8217; dinner, access to the speaker lounge, and a high-quality video recording of your talk. We will also provide 4 nights at the Sheraton San Diego Hotel and round trip economy airfare to San Diego, CA if your employer can&#8217;t sponsor your travel.</p>



<p class="wp-block-paragraph"><strong>Inclusion: </strong>GraphQL Summit welcomes all developers, and we&#8217;re committed to promoting an inclusive environment at our conference. Our speaker lineup will reflect a diverse balance of races, genders, and backgrounds, and we encourage developers belonging to underrepresented groups in tech to apply.</p>



<p class="wp-block-paragraph"><strong>First-time speakers: </strong>If you need help vetting your proposal, please contact us at <a href="mailto:speakers@graphql.com" target="_blank" rel="noreferrer noopener">speakers@graphql.com</a>. We can help coach you on your content and presentation if your talk is accepted.</p>



<p class="wp-block-paragraph"><strong>Code of Conduct: </strong>All attendees, speakers, employees, and volunteers at our conference are required to agree to our Code of Conduct: <a href="https://www.apollographql.com/docs/community/code-of-conduct/" target="_blank" rel="noreferrer noopener">https://www.apollographql.com/docs/community/code-of-conduct/</a></p>



<p class="wp-block-paragraph"><strong>Terms &amp; Conditions: </strong>All attendees are required to accept the <a href="https://summit.graphql.com/2022-apollo-event-registration-agreement.pdf" target="_blank" rel="noreferrer noopener">event terms and conditions</a>. As a speaker, we also ask that you sign a <a href="https://summit.graphql.com/apollo-speaker-consent-and-release.pdf" data-type="URL" data-id="https://summit.graphql.com/apollo-speaker-consent-and-release.pdf" target="_blank" rel="noreferrer noopener">speaker release form</a>.</p>



<p class="wp-block-paragraph">Stay tuned for announcements on the official registration launch, workshop enrollment, hotel discounts, and more! For now, add the event to your calendar, you won’t want to miss this!<br><a href="https://calendar.google.com/calendar/u/0/r/eventedit?dates=20221003T160000Z/20221005T2[%E2%80%A6]%B4%0A%0AStay+tuned+for+details+on+how+to+register.&amp;sf=true" target="_blank" rel="noreferrer noopener">Add to Google Cal &gt;</a><br><a href="https://calendar.google.com/calendar/u/0/r/eventedit?dates=20221003T160000Z/20221005T2[…]%B4%0A%0AStay+tuned+for+details+on+how+to+register.&amp;sf=true" target="_blank" rel="noreferrer noopener">Add to iCal &gt;</a></p>


Mark Your Calendar: GraphQL Summit is Back, Submit your Talk Proposals Now!


<h1 class="wp-block-heading"></h1>



<p class="wp-block-paragraph">Two weeks ago, we hosted a<a href="https://www.apollographql.com/events/tech-talk/from-schema-stitching-to-federation-expedias-journey/"> webinar with Shane Myrick</a>, Senior Software Developer at Expedia Group entitled, “From Schema Stitching to Federation: Expedia’s Journey.” Shane’s talk covered Expedia’s motivation for adopting GraphQL and why they needed a distributed approach for managing their graph.&nbsp;</p>



<p class="wp-block-paragraph">Prior to adopting GraphQL, the Expedia Group faced the challenge of delivering a wide array of travel experiences across a growing number of platforms and clients. Expedia adopted GraphQL as a way to simplify the connection layer between their backend capabilities and front-end clients. The graph provides a single point for clients to connect to and for service teams to expose to clients. Adopting the graph helped Expedia reduce the friction and the number of connections and coordination points across their boundary layers. They also saw benefits from adopting a strongly typed language and the performance benefits of adopting a declarative language that allows clients to request just the data they need. Lastly, they saw benefits in being able to extend and merge common types.</p>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="543" src="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2-1024x543.png" alt="" class="wp-image-22924" srcset="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2-1024x543.png 1024w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2-300x159.png 300w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2-768x407.png 768w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2-1200x637.png 1200w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6433ea44846ed12bb2.png 1276w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<p class="wp-block-paragraph"><em>Building a graph across Expedia’s suite of travel products</em></p>



<p class="wp-block-paragraph"><strong>Scaling their graph with schema stitching</strong></p>



<p class="wp-block-paragraph">As Expedia began to add more capabilities to their graph, they were concerned about it becoming a single point of failure and preventing conflicts as the number of contributors to their graph increased.</p>



<p class="wp-block-paragraph">A core requirement was that they wanted to have one schema and one endpoint for their clients while allowing individual teams to own their services and update those services on their cadence. Expedia started using a gateway as a layer between their GraphQL services and their clients to achieve this. They started off using schema stitching as a way to merge types.</p>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="534" src="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f-1024x534.png" alt="" class="wp-image-22929" srcset="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f-1024x534.png 1024w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f-300x157.png 300w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f-768x401.png 768w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f-1200x626.png 1200w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase6482536c5e195d99f.png 1276w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<p class="wp-block-paragraph"><em>Using a gateway to provide a single endpoint for developers across services</em></p>



<p class="wp-block-paragraph">However, one downside of schema stitching is that Expedia needed to write a significant amount of gateway code to link types across different services. Shane walked through an example of how Expedia wrote code in their gateway to extend a type from their property service to include another type from their reviews service so that clients could query reviews by property. This code enabled the gateway to understand what types existed in different services and merged and extended the schema.</p>



<figure class="wp-block-image size-full"><img loading="lazy" decoding="async" width="894" height="728" src="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase64d41caef3915787fa.png" alt="" class="wp-image-22931" srcset="https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase64d41caef3915787fa.png 894w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase64d41caef3915787fa-300x244.png 300w, https://wp.apollographql.com/wp-content/uploads/2023/11/pngbase64d41caef3915787fa-768x625.png 768w" sizes="auto, (max-width: 894px) 100vw, 894px" /></figure>



<p class="wp-block-paragraph"><em>Writing custom stitching code in the gateway.</em></p>



<p class="wp-block-paragraph"><strong>Challenges emerge with schema stitching</strong></p>



<p class="wp-block-paragraph">As they introduced more stitching code in their gateway, challenges emerged. Their gateway code was becoming more complex, and there was no way to determine the “true schema” without running the gateway. Schema changes required a change to the underlying service and changes to the gateway code and a redeploy of all gateways. This complexity started to cause issues with managing their infrastructure. Additionally, Expedia began to be concerned about preventing conflicts when managing updates to their gateway code.</p>



<p class="wp-block-paragraph"><strong>Moving to Apollo Federation for their graph architecture</strong></p>



<p class="wp-block-paragraph">When Apollo Federation was <a href="https://www.apollographql.com/blog/announcement/apollo-federation-f260cf525d21/">announced</a> in 2019, Expedia began investigating it as an alternative to schema stitching. They liked that<a href="https://www.apollographql.com/docs/federation/"> Federation</a> was published as an open-source <a href="https://www.apollographql.com/docs/federation/federation-spec/">specification</a> and its declarative, directive-based approach to merging types. Federation’s support for all GraphQL servers was compelling as Expedia had invested a lot in Kotlin and even open-sourced their own<a href="https://github.com/ExpediaGroup/graphql-kotlin"> GraphQL Kotlin</a> server.</p>



<p class="wp-block-paragraph">With Federation, Expedia could use GraphQL directives to extend services without writing any custom gateway code. They did have to write resolver code in the underlying service to extend the type, but this was easier than the stitching approach as each team could own the evolution of its service.</p>



<p class="wp-block-paragraph">Expedia saw the following benefits from adopting federation:</p>



<ol class="wp-block-list">
<li>Standardized the implementation as developers could follow the declarative approach to writing schemas.&nbsp;</li>



<li>Able to do offline composition and validation of the schema without spinning up a gateway.&nbsp;</li>



<li>Changed the coordination across teams to review gateway code to have a more strategic conversation about the schema architecture and design. The federated approach made reviewing changes to the schema easier as people reviewed the schema files vs. reviewing many lines of code.</li>
</ol>



<figure class="wp-block-image size-large"><img loading="lazy" decoding="async" width="1024" height="559" src="https://wp.apollographql.com/wp-content/uploads/2021/07/image-11-1024x559.png" alt="image.png" class="wp-image-8113" srcset="https://wp.apollographql.com/wp-content/uploads/2021/07/image-11-1024x559.png 1024w, https://wp.apollographql.com/wp-content/uploads/2021/07/image-11-300x164.png 300w, https://wp.apollographql.com/wp-content/uploads/2021/07/image-11-768x420.png 768w, https://wp.apollographql.com/wp-content/uploads/2021/07/image-11-1200x655.png 1200w, https://wp.apollographql.com/wp-content/uploads/2021/07/image-11.png 1276w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /></figure>



<p class="wp-block-paragraph"> <em>Example of service level resolver code to extend a type across services</em></p>



<p class="wp-block-paragraph"><strong>Making the switch</strong></p>



<p class="wp-block-paragraph">To safely migrate from stitching to Federation, Expedia first worked with the individual service teams to add the necessary federated directives, schemas, and resolver code to their GraphQL servers. Then Expedia set up two separate Apollo servers: one using schema stitching and one set up for Apollo Federation. They set up two variants in Apollo Studio’s schema registry to reflect the two different setups. They used a traffic router in front of the gateways to route traffic between the two graphs, eventually ramping up to a 50/50 split between the stitching and federated gateways.</p>



<p class="wp-block-paragraph"><strong>Benefits from moving from schema stitching to federation</strong></p>



<p class="wp-block-paragraph">By running schema stitching versus Apollo Federation as an A/B test, Expedia was able to quantify the impact of moving to Federation, and they were pleased to see some immediate benefits. First, the federated implementation improved gateway processing latency which improved application performance compared to schema stitching. As Shane explained, “With Federation, we saw a reduced latency compared to schema stitching. Because there was a reduced latency, we were able to reduce our compute quite significantly by about 50%” Lastly, Expedia was able to delete many lines of gateway code which simplified their gateway deployments. Now, they can use almost a stock version of the gateway, making it easy for them to scale the gateway across their many brands since they no longer have much custom code in the gateway.</p>



<p class="wp-block-paragraph"><strong>Key takeaways from moving to Federation</strong></p>



<p class="wp-block-paragraph">After moving to Federation, Shane says Expedia solved key customer and client problems by providing a single endpoint while ensuring that the service teams could independently deploy their changes. They also were able to reduce the potential impact of the gateway going down. They could remove the friction of having to work with multiple code bases to modify the schema. Additionally, they found the declarative approach of the Federation was more straightforward, which made it easier for teams to understand what’s happening and see the overall architecture, which elevated the discussions around the graph and reduced the coordination and complexity of scaling the graph.</p>



<p class="wp-block-paragraph">To learn more about how Expedia’s journey from Schema Stitching to Federation, you can watch Shane’s talk<a href="https://www.apollographql.com/events/tech-talk/from-schema-stitching-to-federation-expedias-journey/"> on-demand</a> now.&nbsp; If you are looking to move from stitching to Apollo Federation, we’ve built a <a href="https://www.apollographql.com/docs/federation/migrating-from-stitching/">migration guide</a>.</p>


Moving from Schema Stitching to Federation: How Expedia improved performance


<p class="wp-block-paragraph">We’re excited to announce that our engineers are hosting the next round of&nbsp;<a href="https://www.eventbrite.com/e/apollo-day-nyc-tickets-59699187878?aff=Blog" target="_blank" rel="noreferrer noopener">Apollo Days in NYC on May 14</a>&nbsp;and&nbsp;<a href="https://www.eventbrite.com/e/apollo-day-san-francisco-tickets-59853663920?aff=Blog" target="_blank" rel="noreferrer noopener">San Francisco on May 22</a>.<br><br>If you’re rolling out GraphQL at your company, whether just getting started or scaling a production deployment, attending an Apollo Day will help you implement practical steps to modernize your apps using GraphQL. You will learn how to:</p>



<ul class="wp-block-list"><li>Plot a strategy to adopt GraphQL on top of existing REST APIs and microservices</li><li>Create a GraphQL microservices architecture by composing schemas into one graph for your company</li><li>Manage local and remote state in React apps seamlessly</li><li>Implement a build/test/deploy workflow that allows product teams to safely make changes to their schema</li><li>Monitor GraphQL-specific performance and errors, integrated with your existing systems like Datadog, PagerDuty, Slack, etc.</li></ul>



<p class="wp-block-paragraph">Come learn best practices from the experts at Apollo and hear technical talks from engineers at companies like&nbsp;<strong>Expedia</strong>&nbsp;and&nbsp;<strong>NY Times</strong>&nbsp;already using it in production.<br><br><strong>SAMPLE AGENDA</strong></p>



<ul class="wp-block-list"><li>12:00pm-1:00pm Check-in &amp; Lunch</li><li>1:00pm-1:30pm Welcome and Keynote</li><li>1:30pm-2:30pm Ask the Experts: A hands-on Apollo platform demo followed by a Q&amp;A session with Apollo engineers</li><li>2:30–3:00pm Coffee break and networking</li><li>3:00–3:30pm The GraphQL developer experience</li><li>3:30–4:30pm Technical deep dives with customers and community members</li><li>4:30–5:00pm Fireside discussion with an Apollo executive and special guest</li><li>5:00–7:00pm Cocktails &amp; Networking</li></ul>



<p class="wp-block-paragraph">Don’t miss this opportunity to hear from and network with some of the leading GraphQL experts in the US.<br><br><strong>Sign up now for&nbsp;</strong><a href="https://www.eventbrite.com/e/apollo-day-nyc-tickets-59699187878?aff=Blog" target="_blank" rel="noreferrer noopener"><strong>NYC on May 14</strong></a><strong>&nbsp;or&nbsp;</strong><a href="https://www.eventbrite.com/e/apollo-day-san-francisco-tickets-59853663920?aff=Blog" target="_blank" rel="noreferrer noopener"><strong>San Francisco on May 22</strong></a></p>


Events

Apollo Summit 2026: Turn Your API Platform Into Your AI Platform

Transforming Capital One’s Data Marketplace with Embedded Governance and a Product Mindset

How N-able Certified 1300% Growth with Apollo

How Delivery Hero Accelerates UX Experiments with Server-Driven UI and Apollo

How Block Scaled Graph Adoption with Apollo and AI

How Wiz Is Redefining Partner Velocity with Apollo MCP

When Every Second Matters: How Athenahealth Protects Patients with Apollo GraphQL

Past posts

Building Trust into AI: How GraphQL Architects Accelerated Developer Delivery

10 Years of GraphQL Celebrated at GraphQLConf 2025

How Indeed’s Bold Bet on Parallel API Platforms Paid Off

Announcing GraphQL Summit by Apollo 2025: Connect Everything. Build Anything.

How Booking.com Orchestrated Their Service Architecture with Apollo Federation

How PlayStation Optimized Performance and Revenue for Their Digital Store with GraphQL Federation

Events

Apollo Summit 2026: Turn Your API Platform Into Your AI Platform

Transforming Capital One’s Data Marketplace with Embedded Governance and a Product Mindset

How N-able Certified 1300% Growth with Apollo

How Delivery Hero Accelerates UX Experiments with Server-Driven UI and Apollo

How Block Scaled Graph Adoption with Apollo and AI

How Wiz Is Redefining Partner Velocity with Apollo MCP

When Every Second Matters: How Athenahealth Protects Patients with Apollo GraphQL

Past posts

Building Trust into AI: How GraphQL Architects Accelerated Developer Delivery

10 Years of GraphQL Celebrated at GraphQLConf 2025

How Indeed’s Bold Bet on Parallel API Platforms Paid Off

Announcing GraphQL Summit by Apollo 2025: Connect Everything. Build Anything.

How Booking.com Orchestrated Their Service Architecture with Apollo Federation

How PlayStation Optimized Performance and Revenue for Their Digital Store with GraphQL Federation

Stay in our orbit