Overview

Lawgic is a Streamlit-based Legal AI Assistant designed to help users interact with their legal documents in a more intuitive way. It uses Google's Gemini 2.5 Flash model for QnA, summarization, and translation, and FAISS vector database for efficient semantic search.

Key Capabilities

• Upload & Analyze: Supports PDF and DOCX legal documents.
• Context-Aware QnA: Ask questions about your documents or general legal queries.
• Custom Summaries: Generate detailed summaries with specific instructions.
• Multilingual Support: Translate legal text into Hindi and Marathi.

How It Works (RAG Pipeline)

1. Indexing: Uploaded documents are chunked and converted into a FAISS vector index.
2. Retrieval: System performs semantic search to find relevant text chunks.
3. Generation: Gemini Model generates the final answer based on retrieved context.

Tech Stack

• Framework: Streamlit (Python)
• AI Model: Google Gemini 2.5 Flash
• Vector DB: FAISS
• Deployment: Render