This is my fork. Prophis was a team project at YHack 2026; the data, the ML, and the statistics were mine. Original repository: sophia-mai/yhack2026. — Nikhil Ramlukan

Prophis

Prophis is a patient-context intelligence tool for public health analysis. It turns fragmented patient history into a readable timeline, connects that case to county-level health patterns, and highlights where earlier preventive action or follow-up could have changed the trajectory.

The current app combines:

• an individual patient timeline
• county-level context from County Health Rankings data
• cohort/context signals such as diabetes similarity and health equity context
• a retrospective prevention review that calls out missed opportunities for earlier action

Run locally

1. Install dependencies

From the repository root:

npm run install:all

2. Configure backend environment

Create a local backend env file from the template:

cp backend/.env.example backend/.env

On Windows PowerShell:

Copy-Item backend/.env.example backend/.env

Then edit backend/.env and set at least one Lava auth token:

LAVA_SECRET_KEY=your_real_lava_secret_key

Or:

LAVA_FORWARD_TOKEN=your_real_lava_forward_token

Optional model overrides:

LAVA_MODEL=openai/gpt-4o-mini
LAVA_MODEL_FAST=openai/gpt-4o-mini
LAVA_MODEL_SUMMARY=openai/gpt-4o-mini
LAVA_MODEL_DEEP=openai/gpt-4o-mini

3. Start frontend + backend

From the repository root:

npm run dev

Frontend: http://localhost:5173

Backend health check: http://localhost:3001/api/health

Notes

• Use npm for this repository.
• The frontend is a Vite React app in frontend/.
• The backend is an Express API in backend/.
• County health data and intervention definitions are stored locally under frontend/public/data/.
• AI routes require either LAVA_SECRET_KEY or LAVA_FORWARD_TOKEN in backend/.env.