Contract-Advisor-RAG: Towards Building A High-Precision Legal Expert LLM APP

Table of Contents

• Project Structure
  • Project overview
  • Tech stack
  • frontend
  • backend
  • notebooks
  • scripts
  • .github
  • Installation guide
  • Contributors
  • License

Project overview

The project goal is to develop an advanced contract AI system for a startup company, creating the first fully autonomous artificial contract lawyer. Using Hybrid LLM technology, the project focuses on building, evaluating, and enhancing a RAG (Retrieval Augmented Generation) system for Contract Q&A. This involves marrying powerful language models with external data sources to provide accurate and context-rich responses. The project includes researching RAG system improvements, building a Q&A pipeline, establishing a RAG evaluation pipeline,optimizing Contract Q&A, and implementing enhancements. Ultimately, it aims to revolutionize the legal industry with an autonomous contract assistant capable of independent contract drafting, review, and negotiation.

Tech Stack

Tech Stack used in this project

• React
• Flask
• Langchain
• chromaDB
• RAGs
• Cohere

frontend

• .frontend/:This dirctory contains the frontend structure.

backend

• .backend/:This dirctory contains the backend structure.

notebooks

• .notebooks/:This directory contains the notebook files.

scripts

• .scripts/: This directory contains the script files.

.github:

• .github/: the folder where github actions and CML workflow is integrated.

Installation guide

Prerequisites:

Clone this repository to your local machine.

https://github.com/Betfsh/Contract-Advisor-RAG.git or
git@github.com:Betfsh/Contract-Advisor-RAG.git

Navigate to the project directory.

cd Contract-Advisor-RAG

Initialize the backend

python app.py

Navigate to the frontend

cd frontend

Contributors

Bethelhem Mebratu

License

• This project is licensed under the MIT License. See the LICENSE file for more information.

Acknowledgements

We would like to thank the Lizzy AI team for their hard work and dedication in developing this project.

• Lizzy AI Website