High Stakes, Real Time: How Athenahealth Scaled GraphQL in Clinical Workflows
summary
Athena uses an ASC workflow as a steel thread to validate GraphQL subscriptions for life‑critical, real‑time updates, wiring React microrends + Apollo Client to an Apollo Server in AWS and the Apollo Router. Resolvers call public REST APIs, the subscription path averages 230 ms with only ~30–40 ms extra network delay and meets the <5s safety target. Adoption scales from 200k to 250M ops in a year, moving to a federated subgraphs architecture (~20 subgraphs), governed by a stewardship committee with MADRs and GraphOS tooling, and exploring Apollo Connectors and the Model Context Protocol.
AI-written summary · about
chapters (21)
- 0:00Athena Health intro
- 1:05Healthcare software mistakes
- 3:17Legacy code challenges
- 5:45Steel thread approach
- 8:15Choosing ASC steel thread
- 9:19Legacy vs GraphQL design
- 13:13GraphQL real-time demo
- 14:59Rapid GraphQL adoption
- 16:22Federated GraphQL at scale
- 17:46GraphQL Stewardship
- 20:38Markdown Architecture Records
- 22:23GraphQL tooling essentials
- 23:27Finding your steel thread
- 24:31Key takeaways
- 25:19GraphQL AI healthcare demo
- 28:57Partners in journey
- 30:01Q&A: MCP setup
- 31:24Q&A: Federated GraphQL
- 32:50Q&A: Democratic governance
- 34:35Q&A: AI hallucinations
- 35:37Q&A: Securing GraphQL
related talks
20mScaling GraphQL on AWS: Production Architecture for High-Volume Data Systems
26mKeynote: Apollo's Journey with GraphQL: Transforming Enterprise APIs for the Future
23mA glimpse of GraphQL" Frontend developers’ life is easy now!"
28mGraphQL at scale with AWS
24mBuilding Reactive GraphQL Apps With Apollo
01h 04mBuilding real-time API's with GraphQL
38mAPI Modernization with GraphQL
58mCrafting compelling real-time web experiences with GraphQL and React