Legal AI — Automated Legal Document Analysis Platform

Full-stack application for contract Q&A and clause understanding.

• client/: Next.js + NextAuth (Google OAuth) UI
• server/: Flask API (Gunicorn) serving ML/NLP endpoints

Quickstart (Docker)

1) Prerequisites

• Docker Desktop or Podman (this repo currently uses podman-compose on macOS)

2) Configure environment

Create a local env file (never commit secrets):

cp client/.env.local.example client/.env

Edit client/.env and set at least:

• NEXTAUTH_URL=http://localhost:3000
• NEXTAUTH_SECRET=... (generate with openssl rand -base64 32)
• GOOGLE_ID=...
• GOOGLE_SECRET=...

Also:

• NEXT_PUBLIC_BACKEND_URL=http://127.0.0.1:5001

3) Build & run

docker compose up -d --build

Open:

• Frontend: http://localhost:3000
• Backend: http://localhost:5001/questionsshort

Local dev (without Docker)

Backend

cd server
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
gunicorn -b 127.0.0.1:5001 app:app --reload

Frontend

cd client
npm install
npm run dev

Key Features

• Contract upload + question selection
• Reading comprehension model (SQuAD-style)
• Paraphrasing (T5-based)
• Sentiment analysis (TextBlob)

Disclaimer

This project is provided for research/education and is not a substitute for professional legal advice.