Legal Assistant with AI

An intelligent legal assistant that helps foreigners understand Polish law through natural language interactions, powered by advanced AI and RAG (Retrieval Augmented Generation) technology.

🎯 Project Goals

This project aims to make Polish legal information more accessible to foreigners by:

• Providing accurate legal information from verified sources
• Offering multi-modal interaction (text, voice, image) //TODO
• Creating a user-friendly interface for legal queries
• Ensuring responses are based on current Polish law

🚀 Features

Data Collection & Processing

• Automated web scraping of legal resources //SEMI-AUTOMATED
• Content filtering and validation
• Structured data storage in Excel/PDF formats
• Exclusion of unofficial or outdated sources

RAG System

• Question-answer retrieval system
• Embeddings-based similarity search
• Context-aware response generation
• Source attribution for transparency

Web Interface

• Real-time chat interface
• Mobile-responsive design
• Conversation history management
• Source reference display

AI Integration

• OpenAI for text generation and embeddings
• Custom RAG implementation for legal contexts
• Fallback mechanisms for reliability

Cloud Deployment

• Frontend hosted on Vercel
• Backend deployed on Google Cloud Run
• Firestore for conversation storage
• Google Cloud Storage for data files

🛠 Technology Stack

Frontend

• React.js
• Vercel for hosting

Backend

• FastAPI
• Google Cloud Run
• Firestore
• Google Cloud Storage

AI/ML

• OpenAI API
• Custom RAG implementation
• Vector embeddings

📋 Project Structure

/
├── frontend/              # React frontend application (Vercel)
│   ├── src/               # React source code
│   └── update-backend-url.js  # Script to update backend URL
│
├── backend/               # FastAPI backend application (Google Cloud Run)
│   ├── app/               # FastAPI application code
│   ├── Dockerfile         # Docker configuration for Cloud Run
│   └── deploy.sh          # Deployment script for Google Cloud
│
└── query/                 # RAG implementation
    ├── prepare_rag.py     # RAG system implementation
    └── *.xlsx             # Legal Q&A data files

🚀 Deployment

Backend Deployment (Google Cloud Run)

1. Install the Google Cloud SDK:

   curl https://sdk.cloud.google.com | bash
   gcloud init
   

2. Set your OpenAI API key:

   export OPENAI_API_KEY=your-openai-api-key
   

3. Run the deployment script:

   cd backend
   chmod +x deploy.sh
   ./deploy.sh
   

4. Note the Cloud Run URL from the output.

Frontend Deployment (Vercel)

1. Update the backend URL in the frontend:

   cd frontend
   node update-backend-url.js YOUR_CLOUD_RUN_URL
   

2. Deploy to Vercel:

   vercel --prod
   

For detailed deployment instructions, see the backend/README.md and DEPLOYMENT.md files.

🔜 Roadmap

1. Phase 1: Data Collection & Processing ✅

   • Enhance web scraping reliability
   • Improve content filtering
   • Expand data sources

2. Phase 2: RAG System Enhancement ✅

   • Fine-tune embeddings for legal context
   • Implement better answer generation

3. Phase 3: Web Interface ✅

   • Develop React frontend
   • Create Python backend

4. Phase 4: Cloud Deployment ✅

   • Frontend on Vercel
   • Backend on Google Cloud Run
   • Firestore for data storage

5. Phase 5: Future Enhancements

   • Add voice input/output
   • Enable image-based queries
   • Implement multi-language support