Real estate price prediction app.
Pre-requisites: Docker.
Trains the model, builds the frontend React app and runs everything: make run.
Smoke test the API with make test-api.
There are two applications: Python Backend and React Frontend. Frontend also acts as a reverse proxy for the /api route.
The backend app has two sub-modules: the JSON REST API that's served by the FastAPI, and an ML model training script. The training script is very coupled to the sample dataset. The model is trained during docker build. See Backend Dockerfile for details.
Frontend is a simple Vite+React+TypeScript app, built during docker build (see Frontend Dockerfile for details) served by an nginx container (see nginx config). Nginx also forwards all requests to /api route down to backend container.
- Fix local dev environment to be able to submit form from frontend to backend container without messing with production setup.