Speed Without Sacrifice: How Wayfair Transforms DevEx with AI and MCP
summary
Wayfair embeds Apollo MCP and off‑the‑shelf LLMs to build AI schema docs, RAG‑powered schema reviews, and AI schema mocks that actually run in Apollo Studio. At 200+ subgraphs and 3B daily requests, an AI doc service updates 1,600 fields, achieves 92% PR acceptance and cuts undiscoverable fields from 70% to 50%. RAG plus prior reviews drops time‑to‑first‑comment from 3.5 days to instant and approval from 6 to 4.5 days (25%); mocks speed front‑end time‑to‑market 10–20% with Zod validation and post‑processing.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:24Speed and quality in GraphQL
- 3:33Schema docs quality problem
- 6:45AI schema docs and implementation
- 10:01AI docs impact on discoverability
- 12:16AI schema reviews at design stage
- 14:06Demo: AI schema review in Apollo Studio
- 17:13Building AI schema reviews with RAG
- 21:47AI schema reviews results
- 23:13AI schema mocks for testing
- 26:37AI schema mocks in workflows
- 31:13Key takeaways
related talks
28mSpeed Without Sacrifice: How Wayfair Transforms DevEx With AI and MCP
27mPanel: APIs for AIs
21mSimplifying MCP Tool Sprawl With GraphQL
26mAsk, Don't Fetch: Why GraphQL + MCP Is the Native Language of AI Agents
30mPrompt to Performance Speeding Up Database Administration with AI and Claude Code
34mYou say you've got an "AI SDLC"
31mAccelerate AWS Well-Architected reviews with Generative AI
13mFaster AI time-to-value when planning