Workshop: Demand-Driven Schema Design
summary
Workshop teaches reverse-engineering a client dream query into a practical GraphQL schema, using concrete tools and patterns: quick SDL mocking with Apollo Server, composing subgraphs with Rover Dev and a supergraph.yaml, and performance gains from caching product subgraphs. Advice is specific: don’t copy REST payloads or expose transient ID fields, model entities (including compound/time-based keys), use the nested mutation pattern for entity operations, design explicit error shapes and use extensions for observability; live demo mixes mocks and federation composition.
AI-written summary · about
chapters (9)
- 0:00Workshop introduction
- 4:51Common schema mistakes
- 5:13Designing for client needs
- 18:39Schema governance process
- 23:13Backend perspective on schema
- 32:14Schema mocking tools
- 43:53Federated architecture patterns
- 1:06:25Time-based graph design
- 1:19:02Workshop recap and Q&A
related talks
28mTeach Yourself GraphQL in 2026: An Anti-blueprint
47mSchema Design in the Age of AI: Pragmatic Patterns to Evolve and Ship Fast
01h 14mWorkshop: GraphQL on Client and Server — From Idea to Prototype (Part 2)
33mDesigning a GraphQL Schema
49mGraphQL Schema Design — Building Schemas Correctly (version 2)
01h 12mWorkshop: GraphQL on Client and Server — From Idea to Prototype (Part 1)
57mElm GraphQL Workshop
30mSchema Design Patterns: Leveraging Existing REST APIs for Rapid GraphQL