LegalSmart Chatbot

LegalSmart Chatbot is a Retrieval-Augmented Generation (RAG) legal assistant focused on Indian law. It answers user questions using context retrieved from the Indian Penal Code (IPC), Bharatiya Nyaya Sanhita (BNS), and the Constitution of India.

The project is implemented with Streamlit, LangChain, FAISS, Hugging Face embeddings, and Together AI-hosted LLMs.

Research Publication

This project is part of our published research work:

• Springer (book chapter): https://link.springer.com/chapter/10.1007/978-3-032-12827-0_7

Features

• Grounded legal Q&A over IPC, BNS, and Constitutional text.
• Conversational memory for multi-turn interactions.
• Retrieval pipeline powered by FAISS vector search.
• Streamlit-based chat interface for easy deployment.
• Prompting strategy designed for both non-legal and legal-professional audiences.

Tech Stack

• App/UI: Streamlit
• LLM Orchestration: LangChain
• Vector Store: FAISS
• Embeddings: nomic-ai/nomic-embed-text-v1 (Hugging Face)
• LLM Provider: Together AI

Project Structure

.
├── app.py                  # Streamlit chat app
├── Ingest.py               # Document ingestion + vector indexing
├── requirements.txt        # Python dependencies
├── data/                   # Source legal PDFs (IPC, BNS, Constitution)
└── law_vector_db/          # Generated FAISS index files

Getting Started

1) Clone the repository

git clone https://github.com/srujankothuri/LegalSmart-Chatbot.git
cd LegalSmart-Chatbot

2) Install dependencies

pip install -r requirements.txt

3) Build the vector database

Run ingestion first (recommended on a machine with sufficient RAM/CPU):

python Ingest.py

This generates/updates the law_vector_db directory used at runtime.

4) Configure API key

Set your Together AI API key as an environment variable:

export TOGETHER_API_KEY="your_together_api_key"

If deploying on Streamlit Cloud/Hugging Face Spaces, store TOGETHER_API_KEY in the platform's secrets settings.

5) Run the app

streamlit run app.py

Then open the local URL shown in your terminal (typically http://localhost:8501).

Notes

• This tool is for educational and informational purposes.
• Responses may not be a substitute for professional legal advice.
• Always verify critical legal decisions with a qualified legal practitioner.

Contributing

Issues and pull requests are welcome. If you find bugs or want improvements, please open an issue first to discuss scope.

Contact

For feedback or questions, open an issue:

• https://github.com/srujankothuri/LegalSmart-Chatbot/issues