This is a submission for the Open Source AI Challenge with pgai and Ollama
What I Built
I developed an AI-powered application using a Retrieval-Augmented Generation (RAG) system to improve accessibility to the Ivorian Labor Code. This legal document is over 500 pages long, often complex, and challenging for the general public to navigate. My solution makes it easier to quickly find relevant information and reformulates the content to improve understanding. This project stemmed from my own difficulty in navigating the Labor Code and my desire to provide a practical, accessible solution for everyone.
The application leverages powerful open-source tools for vector management and text generation, including pgvector making it possible to implement a RAG system suited to a large text corpus. My goal is to make Ivorian legislation more accessible, save users valuable time, and enhance the experience of those engaging with dense legal texts.
Demo
- ngrok demo link app
Tools Used
To build this RAG system, I used several specific open-source tools:
- pgvector: Used to store and manage vectors of the Labor Code text. This allows for quick access to relevant passages based on the user's query.
- ollama Model : llama3.2
- streamlit for chat app
Final Thoughts
Developing this application has been an enriching experience. Not only did I gain deeper insights into open-source vector management tools, but I also explored how AI can be used to democratize access to complex legal information. I hope this project will contribute to a better understanding of the Ivorian Labor Code and inspire similar solutions for other legal documents.
This project is submitted for the following categories: Open-source Models from Ollama
Top comments (0)