Dr. Aigent
Multi-model clinical AI that turns doctors' notes into structured SOAP reports.
- Backend and AI engineer
- June 2025 to present
- Python, FastAPI, Multi-model LLM routing
The problem
Doctors spend real time turning messy consultation notes into structured reports. The goal with Dr. Aigent was to take a doctor's notes and produce a clean, structured SOAP report, reliably enough to trust in a working day, not just in a demo.
What I built
Dr. Aigent is a multi-model clinical AI that turns doctors' notes into structured SOAP reports. I built the Python and FastAPI service behind it and the model layer that does the work.
The hard parts
Routing across models, with fallback
The reliability comes from routing across GPT, Claude, and Gemini with automatic, cost-aware fallback. If one model is unavailable or too expensive for a given request, another picks it up, so the tool keeps producing reports in production instead of failing when a single provider has a bad day.
Where it is now
Dr. Aigent is in production, as one of three connected platforms I built for the same client.