Home
Softono
legal-tech-chat

legal-tech-chat

Open source MIT Jupyter Notebook
Website GitHub
158
Stars
34
Forks
0
Issues
4
Watchers
1 year
Last Commit
AI Agents Data Visualisation

About legal-tech-chat

Extract structured data from CUAD contracts using LangChain, build a knowledge graph, and query insights through a LangGraph agent - transforming legal agreements into actionable intelligence.

Platforms

Web Self-hosted

Languages

Jupyter Notebook

Links

Website Source Code
t
Published by
tomasonjo-labs
Visit View Profile

README.md

View on GitHub

Agentic GraphRAG on commercial contracts

This repository processes the Contract Understanding Atticus Dataset (CUAD) using LangChain to extract structured data from legal contracts and build a comprehensive knowledge graph. The system includes a LangGraph-powered agent for intelligent querying and analysis of contract information.

See blog for more: https://towardsdatascience.com/agentic-graphrag-for-commercial-contracts/

đź“„ About CUAD

The Contract Understanding Atticus Dataset (CUAD) consists of 500 contracts with annotations for 41 legal clauses. This dataset provides a rich source of legal text for information extraction and analysis.

🚀 Features

  • Contract Processing: Extract structured data from 500 CUAD contracts using LangChain
  • Knowledge Graph Construction: Build a comprehensive graph database of contract relationships and clauses
  • Intelligent Query System: Utilize LangGraph agents to query and analyze the contract database

🛠️ Setup and Installation

  1. Clone the repository
  2. Copy the environment file:
    copy .env.example to .env
  3. Start the application:
    docker-compose up