LegisAI - AI-Powered Legal Assistant

A production-ready legal AI assistant with intelligent document analysis, multi-agent orchestration, clause risk detection, and hybrid search capabilities.

🚀 Features

Core Capabilities

• ✅ Legal Research - Hybrid search (FAISS + BM25) with 1,200+ contract clauses
• ✅ Document Drafting - AI-powered contract generation with streaming responses
• ✅ Compliance Checking - GDPR/CCPA and jurisdiction-aware analysis
• ✅ Clause Analysis - 4-tier risk scoring (low/medium/high/critical)
• ✅ Risk Visualization - Interactive D3.js heatmap for clause-level risks
• ✅ Document Upload - PDF processing and analysis
• ✅ Streaming Responses - Real-time word-by-word generation

Advanced AI Agents

• 🔍 Summarization Agent - Extract headnotes, ratio decidendi, obiter dicta
• 🎯 Precedent Reasoning Agent - Cross-case alignment, outdated precedent detection
• 📊 Knowledge Graph Builder - Auto-builds graph of precedents and statutes
• 📝 Redlining & Comparison Agent - ML-based clause alignment, risk-focused changes
• ⚙️ Clause Generation Agent - Detect missing clauses, generate standard clauses
• 📋 Regulatory Monitoring Agent - Track regulations, flag new requirements
• 🎲 Monte Carlo Risk Agent - What-if scenario simulations
• 🔄 Multi-Agent Orchestration - LangGraph for complex workflows with state management

Technical Features

• 🔍 FAISS Vector Store - Semantic search with sentence transformers
• 🎯 Hybrid Retrieval - Combines semantic (FAISS) + keyword (BM25) search
• 📚 Real Legal Data - 1,200+ contract clauses from CUAD dataset
• ⚖️ Jurisdiction-Aware - US, EU, UK legal rules
• 🔒 Cross-Consistency Checking - Multiple agents validate same clause
• 🌐 Graph-Based Workflows - LangGraph orchestration with conditional routing

📊 Project Status

Overall Completion: ~95% of core features

• ✅ 16 Major Features Fully Implemented
• ⚠️ 5 Features Partially Implemented
• ❌ 4 Features Not Yet Implemented

See FEATURE_COMPARISON.md for detailed status.

🚀 Quick Start

Prerequisites

For Docker:

• Docker Desktop installed and running
• At least 8GB RAM
• ~15GB free disk space

For Local Development:

• Python 3.11+ installed
• Node.js 18+ installed
• Ollama installed and running locally

Start the Application

Option 1: Docker (Recommended)

# Start Backend with Docker
docker-backend.bat

# Start Frontend with Docker (in another terminal)
docker-frontend.bat

Option 2: Local Development

# Start Backend locally
backend.bat

# Start Frontend locally (in another terminal)
frontend.bat

Access the Application

• Frontend: http://localhost:5173
• Backend API: http://localhost:8000
• API Documentation: http://localhost:8000/docs

For detailed instructions, see QUICK_START.md.

📁 Project Structure

LegisAI/
├── backend/
│   ├── main.py                    # FastAPI application (47 endpoints)
│   ├── requirements.txt           # Python dependencies
│   ├── services/                  # 14 AI agent services
│   │   ├── vector_store.py       # FAISS integration
│   │   ├── hybrid_retriever.py   # BM25 + FAISS
│   │   ├── clause_analyzer.py    # Risk detection
│   │   ├── summarization_agent.py
│   │   ├── precedent_reasoning_agent.py
│   │   ├── knowledge_graph_builder.py
│   │   ├── redlining_comparison_agent.py
│   │   ├── clause_generation_agent.py
│   │   ├── regulatory_monitoring_agent.py
│   │   ├── monte_carlo_risk_agent.py
│   │   ├── langgraph_orchestrator.py  # LangGraph workflows
│   │   └── courtlistener_api.py
│   ├── data/                      # Vector store data
│   └── uploads/                   # User uploads
├── frontend/
│   ├── src/
│   │   ├── pages/                # React pages
│   │   ├── components/           # UI components
│   │   └── utils/                # Utilities
│   └── package.json
├── backend.bat                    # Start backend locally
├── docker-backend.bat             # Start backend with Docker
├── frontend.bat                   # Start frontend locally
├── docker-frontend.bat           # Start frontend with Docker
└── docker-compose.yml             # Full stack Docker setup

🛠️ Installation

Quick Start Scripts

Windows Batch Files:

• backend.bat - Start backend locally
• docker-backend.bat - Start backend with Docker
• frontend.bat - Start frontend locally
• docker-frontend.bat - Start frontend with Docker

Usage:

# Docker (Recommended)
docker-backend.bat    # Terminal 1
docker-frontend.bat   # Terminal 2

# Local Development
backend.bat           # Terminal 1
frontend.bat          # Terminal 2

📝 API Documentation

Full interactive API docs available at http://localhost:8000/docs when backend is running.

Key Endpoints

Research & Retrieval:

• POST /api/research - Legal research with hybrid search
• POST /api/summarize/headnotes - Extract case headnotes
• POST /api/summarize/ratio-decidendi - Extract legal reasoning
• POST /api/precedent/align - Find supporting/weakening cases
• POST /api/knowledge-graph/build - Build knowledge graph

Drafting & Contract Intelligence:

• POST /api/draft - Generate legal documents
• POST /api/clauses/analyze - Clause risk analysis
• POST /api/redlining/align - Compare two contracts
• POST /api/clauses/generate - Generate missing clauses
• POST /api/clauses/detect-missing - Detect missing standard clauses

Compliance & Risk:

• POST /api/compliance/check - Compliance analysis
• POST /api/regulatory/crawl - Monitor regulations
• POST /api/risk/monte-carlo - Risk scenario simulations

Orchestration:

• POST /api/langgraph/comprehensive-analysis - Multi-agent comprehensive workflow
• POST /api/langgraph/cross-consistency-check - Consistency validation workflow
• POST /api/langgraph/research-and-draft - Research → draft → review workflow
• GET /api/langgraph/status - Orchestrator status

Document Management:

• POST /api/upload - Upload documents
• GET /api/analyze/{file_id} - Analyze uploaded document

📊 Data Sources

Currently Integrated

• CUAD Dataset - 1,200+ contract clauses imported and indexed
• Vector Store - 1,267+ total documents (FAISS + BM25 indexed)
• CourtListener API - Real-time case law search (optional, requires API key)
• Regulations - GDPR, CCPA, US Code sections

🐳 Docker Configuration

Environment Variables

Create backend/.env:

# Ollama Configuration
OLLAMA_HOST=http://ollama:11434  # Docker
# OLLAMA_HOST=http://127.0.0.1:11434  # Local

# CourtListener API (optional)
COURTLISTENER_API_TOKEN=your_api_token_here

# Application
PORT=8000
ENVIRONMENT=production

Docker Commands

# Start services
docker compose up -d

# View logs
docker compose logs -f

# Stop services
docker compose down

# Check status
docker compose ps

📚 Documentation

• README.md - This file (main documentation)
• FEATURE_COMPARISON.md - Complete feature status and project progress
• QUICK_START.md - Quick setup guide
• CONTRIBUTING.md - Contribution guidelines
• COMPLEXITY_ANALYSIS.md - Code complexity metrics
• HOW_ORCHESTRATION_WORKS.md - Multi-agent orchestration guide
• LANGGRAPH_IMPLEMENTATION.md - LangGraph setup guide
• HOW_ORCHESTRATION_WORKS.md - Multi-agent orchestration guide

🏗️ Architecture

Backend

• Framework: FastAPI (Python)
• LLM: Ollama (local LLM hosting)
• Vector Store: FAISS (semantic search)
• Search: Hybrid (FAISS + BM25)
• Orchestration: LangGraph (graph-based workflows)
• API: RESTful with 47 endpoints

Frontend

• Framework: React + TypeScript
• Build Tool: Vite
• Styling: TailwindCSS
• Visualization: D3.js (risk heatmap)

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

See CONTRIBUTING.md for details.

📄 License

MIT License - See LICENSE file for details

🆘 Support

For issues or questions:

• Create an issue on GitHub
• Check the documentation
• Review QUICK_START.md for troubleshooting

🎉 Acknowledgments

• Ollama for local LLM hosting
• FAISS for vector search
• LangGraph for multi-agent orchestration
• FastAPI for the backend framework
• React + Vite for the frontend

---

Built with ❤️ for the legal community

Version: 4.0.0
Status: Production Ready (95% Complete)