AI, FHIR & the future of healthcare data
summary
Conversation argues that standards remain essential while AI augments them: FHIR continues as the best input/output format, provenance and confidence must be tracked, and PIQI is a starting framework for data quality. Predictive LLMs (trained on public text including Reddit) need structured inputs so b.well-style data refinery pipelines clean and normalize messy source records; built lab-mapping routines already scale to millions of patient records. Open licensing of FHIR and agents plus emerging agent orchestration drive near-term change.
AI-written summary · about
chapters (9)
- 0:00Welcome and introductions
- 1:51AI and structured data exchange
- 6:54FHIR and AI compatibility
- 13:49Data quality as AI bottleneck
- 18:18FHIR licensing and AI impact
- 22:18Terminology standards and AI
- 27:38AI accountability in healthcare
- 31:40Future of AI in healthcare workflows
- 35:13Standards evolution with AI
related talks
47mThe Future of Healthcare and AI
33mBridging large-scale FHIR to AI in production
52mPanel: AI-Enhanced Healthcare: Future Trends
25mYou put what, where? What FHIR should be doing about AI/ML
22mAI and ML for Centralized Patient Data Repositories
28mThe Future of Healthcare – AI’s Promise and the Challenges Ahead
15mTransforming Healthcare with BI & AI
21mGen AI in Health: What's New in 2026?