Your next million API consumers won't be developers. They'll be AI agents. And they don't read documentation, parse hypermedia links, or guess which of your 200 REST endpoints returns the data they need.
\nThis talk examines what happens when autonomous AI agents become the primary consumers of your API layer. Drawing on real data from Singapore's public government APIs, I'll show how REST responses waste 30–60% of an agent's token budget on structural overhead, and how a typed, self-describing schema changes the equation entirely.
\nWe'll walk through the three properties that make an API truly agent-native: discoverability, precision, and composability. We'll look at what it would take to unify API estates like Singapore's 3,000+ government APIs across 75+ agencies into a single, self-describing surface. A pattern Gartner expects 30% of enterprises to adopt by 2027.
\nYou'll leave with a framework for what makes an API truly agent-native, why GraphQL's type system and federation model get you there, and how to start without a rewrite.
\n", @@ -49,8 +49,8 @@ export const londonSessions: EventSession[] = [ id: 4002, uuid: "52b3a4d6-b5ac-40d8-bc1a-c4bb5cf2bb2f", title: "Explaining GraphQL Federation to my non-technical spouse", - start: "2026-10-01T10:00:00+01:00", - end: "2026-10-01T10:25:00+01:00", + start: "2026-10-01T10:45:00+01:00", + end: "2026-10-01T11:10:00+01:00", tags: ["GraphQL", "Federation"], description: "In this talk, I will explain the core concepts of GraphQL Federation as accessibly as possible. The goal is that you could show this talk to anyone non-technical (such as my wife!) and they'd be able to understand and follow along. Some example topics include what subgraphs and supergraphs are, what entities are, what composition directives are, etc. It will be a fun, interactive (with audience participation!) experience.
\n", @@ -75,8 +75,8 @@ export const londonSessions: EventSession[] = [ id: 4003, uuid: "4178e7d1-37fd-4f70-8519-1aab3b34a48b", title: "Teach Your AI Agent GraphQL", - start: "2026-10-01T10:30:00+01:00", - end: "2026-10-01T10:55:00+01:00", + start: "2026-10-01T11:30:00+01:00", + end: "2026-10-01T11:55:00+01:00", tags: ["GraphQL", "AI Agents"], description: "You can hand your GraphQL work to an AI agent today, but it tends to make the same mistakes developers have made for years: schemas that are nullable everywhere, anonymous queries, and code that ignores partial errors. Most of the time the model isn't the problem. It just doesn't know your API or your team's conventions, so it guesses.
\nAgent Skills are a lightweight, open format for giving an agent that missing context. In this talk, I'll walk through real examples from Apollo's open-source skills and show how a few Markdown files can nudge an agent toward efficient operations that respect your conventions and intentional schema design that scales with your team, instead of leaving it to guess.
\nYou don't need to be an expert in AI or GraphQL to follow along. If you've ever used a coding assistant and wished it understood GraphQL better, this talk is for you.
\n", @@ -99,8 +99,8 @@ export const londonSessions: EventSession[] = [ uuid: "0d452c7b-b0a5-4021-ae49-46f3d75da007", title: "Server Assisted Accessibility: Embedding UI Semantics Into Your GraphQL Schema + CI", - start: "2026-10-01T11:30:00+01:00", - end: "2026-10-01T11:55:00+01:00", + start: "2026-10-01T11:55:00+01:00", + end: "2026-10-01T12:20:00+01:00", tags: ["GraphQL", "Accessibility"], description: "Accessibility is often an afterthought or is left until the last milestone before a feature ships. What if we shifted that left? This talk showcases a sandbox idea: embedding accessibility metadata directly into GraphQL schemas using custom directives and code generation. This enables clients (Android, iOS, and web) to get accessible defaults from day one.
\nWe'll cover the directive design, how to expose metadata in generated types, and how frontends like Jetpack Compose or SwiftUI consume it. We'll walk through a practical, low friction approach: adding CI friendly validation that catches common contract regressions before changes ship, such as missing required metadata, invalid values, and template drift.
\nYou'll leave with schema snippets you can copy & paste, strategies to introduce this in your own APIs, and a blueprint for making accessibility part of your developer experience. A repository with working code examples will be provided.
\n", @@ -125,8 +125,8 @@ export const londonSessions: EventSession[] = [ id: 4005, uuid: "47efc864-717c-4258-90f3-dc5ed0839b82", title: "Houston, We Have a Partial Response: GraphQL Error Handling", - start: "2026-10-01T12:00:00+01:00", - end: "2026-10-01T12:25:00+01:00", + start: "2026-10-01T12:20:00+01:00", + end: "2026-10-01T12:45:00+01:00", tags: ["GraphQL", "Error Handling"], description: "Your REST instincts are lying to you. In GraphQL the status code is almost always 200, errors travel in an array alongside your data, and one null in the wrong place can wipe out the entire response. Welcome to the partial response: a successful failure.
\nWe'll work on problems like mission control: how errors propagate and bubble, top-level errors versus errors as data, union result types, and what clients should do when half the payload burns up on re-entry. You'll leave with a clear framework for which errors belong in your schema and which belong in the errors array.
\nFailure is not an option. It's a field.
\n", @@ -148,8 +148,8 @@ export const londonSessions: EventSession[] = [ id: 4006, uuid: "4d529f8a-1208-4e7d-a9d3-ee2694581cf0", title: "GraphQL with AWS AppSync - Building a Product API at HEMA", - start: "2026-10-01T12:30:00+01:00", - end: "2026-10-01T12:55:00+01:00", + start: "2026-10-01T12:45:00+01:00", + end: "2026-10-01T13:10:00+01:00", tags: ["GraphQL"], description: 'I will talk about "pioneer phase" of GraphQL adoption at HEMA and using AWS AppSync for the composable commerce migration, where I will share the raw and honest lessons learned when introducing a new tech stack to a 100-year-old Dutch icon.
\n',