The Health Data Agent is a healthcare AI project focused on clinical intelligence, interoperability, and AI governance for sensitive healthcare data. It combines practical AI engineering with PHI-safe controls that make clinical workflows easier to review, audit, and trust.
AI governance and security controls for PHI-safe workflows, request guardrails, and audit-aware clinical AI access.
Ingest legacy HL7 v2 messages and convert them to FHIR R4, using AI to analyze clinical notes and extract key insights.
A practical AI interface for asking patient-data questions, reviewing insights, and exploring clinical risk signals.
Governance & Security Focus
The Warden is the core security and governance layer: it keeps PHI out of model exposure, applies request-level guardrails, and records audit-aware governance events. The broader system is built with Python/FastAPI, AWS Bedrock, typed Pydantic validation, HL7/FHIR workflows, and ChromaDB for semantic memory.