Bodh
Medical AI platform with physician panel orchestration
Bodh is an open-source medical AI platform that orchestrates a virtual physician panel on the Microsoft Agent Framework. The system is inspired by Microsoft's MAI-DxO and SD-Bench, implementing role-specialized agents for intake, questioning, test planning, diagnostician analysis, and reasoning verification.
The platform is fully aware of healthcare standards (FHIR R4 for structured medical data, HL7 v2 for clinical messaging) and implements cost-aware diagnostic budget enforcement to optimize care delivery while managing expenses.
Bodh follows a multi-agent sequential diagnosis workflow. Each patient case progresses through a structured pipeline: the intake agent collects and normalizes patient data, the questioning agent gathers clinical history through targeted follow-ups, the test planner recommends appropriate diagnostics, the diagnostician agents analyze results in parallel, and the reasoning verifier validates the final assessment for logical consistency.
All patient data handling is designed with HIPAA-compliant safeguards. Protected health information is encrypted at rest and in transit, access is scoped per agent role, and audit logs capture every data access event for compliance review. The platform enforces clinical decision support with human-in-the-loop oversight, ensuring that AI-generated insights serve as decision aids rather than autonomous actions.
Key Highlights
- Virtual physician panel with 6+ role-specialized diagnostic agents
- FHIR R4 and HL7 v2 standards compliance for healthcare interoperability
- Cost-aware diagnostic budget enforcement
- Inspired by Microsoft's MAI-DxO research
- Deployment-ready on Kubernetes with OpenTelemetry observability