Personal Project · AI / Legal Tech
Labour Code Assistant
A chatbot that uses Retrieval-Augmented Generation to answer questions about the Chadian Labour Code. You ask a question in plain language, it finds the relevant articles and gives you a cited answer.
Why this exists
The Chadian Labour Code is a dense legal document, and most people who need to reference it (workers, small business owners, HR staff) don't have easy access to a lawyer. I wanted to see if a RAG pipeline could make this kind of document usable for a non-expert by letting them ask questions in plain language and getting answers that point to the exact article.
How It Works
From PDF to answer in five steps
1
Load
Parse the Labour Code PDF into raw text
2
Split
Break into smaller chunks for comparison
3
Embed
Convert chunks to vectors in ChromaDB
4
Retrieve
Semantic search for relevant chunks
5
Generate
LLM writes an answer citing specific articles