Bmiami988/Uganda-legal-AI

A production-grade RAG 3.0 legal awareness chatbot built on Ugandan legislation. Combines FAISS vector search, cross-encoder reranking, and Groq LLM inference to deliver fast, grounded, and citation-aware legal answers. **UGANDA LEGAL AI** UgandaLegalAI is a modern Retrieval-Augmented Generation (RAG 3.0) system built to improve access to Ugandan law. The system ingests official legislation PDFs, builds a persistent FAISS vector index, applies cross-encoder reranking for precision retrieval, and uses Groq-powered large language models to generate grounded, detailed, and context-aware responses. Designed for community legal awareness, NGOs, and civic-tech deployment, the system emphasizes: Grounded answers using official Acts Low hallucination risk through strict context control Fast inference via Groq Persistent vector indexing (no re-embedding required) Modern, responsive chatbot UI built with Streamlit This project demonstrates how advanced RAG architectures can be used for legal access and public empowerment in emerging markets.