AI and ML for Centralized Patient Data Repositories
summary
Architecture description focuses on concrete claims not promises: a 38% operational cost reduction and 52% improved data accessibility from cloud‑native deployment, storage choices like Postgraql for structured data and MongoDB for unstructured with 99.99% uptime, an implemented FHIR API layer handling 850,000 daily API requests and security using TLS 1.3 and RBAC+ABAC, and a real‑time CDC pipeline processing around 500 changes per second with sub‑150 ms latencies.
AI-written summary · about
chapters (12)
- 0:00Introduction to Centralized Data Repositories
- 0:21Challenges in Healthcare Data Management
- 1:29Microsoft Architecture and Cloud Approach
- 3:19Data Storage and Integration Layers
- 5:53Security Measures in Healthcare Systems
- 6:59Designing Data Models for Healthcare
- 7:58Data Matching and Automation
- 9:50FHIR API Layer Implementation
- 11:08Real-Time Synchronization Architecture
- 16:15Performance Optimization in Healthcare Systems
- 19:43Monitoring and Operations
- 21:16Conclusion and Final Thoughts
related talks
39mAI, FHIR & the future of healthcare data
22mBuilding Predictive Analytics Data Warehouses in Healthcare
48mAI for Healthcare
08mRevolutionizing Healthcare with Big Data and ML
48mAnalytics and AI on FHIR
33mBridging large-scale FHIR to AI in production
36mAI in healthcare - what is hype and what is real?
09mHow AI is changing the healthcare system