A web application that processes PDF reports from baropodometric (foot pressure) tests, extracts clinical metrics, and generates interpretations using OpenAI's Assistant API.
- Uploads baropodometry PDF reports (neutral stance, closed eyes, cotton rolls)
- Extracts structured metrics via a 3-stage OpenAI assistant pipeline
- Compares baseline vs. intervention (with/without device) states
- Produces clinical interpretation text for practitioners
- Next.js + TypeScript — web frontend and API routes
- OpenAI Assistant API (
gpt-4o-mini,code_interpreter,file_search) — extraction and analysis - Netlify / Vercel — deployment (see
netlify.toml,vercel.json)
npm install
cp env.example .env.local # fill in your OpenAI credentials
npm run devOpen http://localhost:3000.
SETUP.md— environment setupSETUP_LOCAL_DEPLOYMENT.md— local production deploymentSYSTEM_WORKFLOW.md— analysis pipeline and assistant architecture
MIT — see LICENSE.