Supergraph Saving Lives
summary
Startup moves a monolith to Apollo Federation Supergraph and ships a request engine (2.5 months planning, ~3 weeks build) and a survey/assessment engine (2.5 months planning, ~1 week build) that uses factor analysis to cut question counts and models surveys as a directed graph for O(1) lookups. A permission evaluator caches a map with subscriptions, cutting authorization checks from ~250–600 ms to ~80 ms. Clinical validation, governance, and security run in-line and enterprise adoption improves via embedded clinician workshops.
AI-written summary · about
chapters (9)
- 0:00The founder's journey
- 1:34Engineering and leadership
- 3:21Overcoming personal adversity
- 5:13Humanizing technology
- 6:49GraphQL as a strategic tool
- 11:26Modeling complex data
- 16:25Customer-centric development
- 20:27Empowering engineering teams
- 24:31Future initiatives and AI
related talks
30mDeep Learnings Migrating legacy services to Federated GraphQL
34mGraphQL as a Data Mesh Access Layer in Global Banking
21mSafely Merging Subgraphs in a Distributed World
27mGraphQL Meets LLMs & Agents: Building Production AI at Starbucks Scale
38mGraphQL Federation in the Wild
33mThe Future of Efficiency Is Here: Add Planning to Your Schema!
07mExperience-Oriented Graphs with Batching
48mAPI Optimization Tale: Monitor, Fix and Deploy (on Friday)