⚖️ LawgicAI-Chatbot

💡 A Streamlit-based legal chatbot powered by RAG architecture, built to deliver hallucination-free, document-grounded answers — unlike generic LLMs.

---

📸 Demo

---

🚀 Features

• 🧠 RAG (Retrieval-Augmented Generation) backed memory
• 🗃️ Custom vectorstore using FAISS for fast retrieval
• 🛡️ Avoids hallucinations — answers are based on your uploaded documents
• 🧩 Modular code (create_memory.py, connect_memory.py, chatbot.py)
• ⚡ Instant and interactive UI with Streamlit
• 📁 Clean folder structure with data/ and vectorstore/ separation

---

🧠 Architecture

📁 data/
   └── your uploaded PDFs
📁 vectorstore/db_faiss/
   └── vector index stored here
📄 create_memory.py
   └── Converts PDFs to chunks → embeds with LLM → stores vectors
📄 connect_memory.py
   └── Loads vectorstore and connects it to LLM
📄 chatbot.py
   └── Streamlit UI to chat with your knowledge base

---

💡 Why LawgicAI is Better

Feature	LawgicAI	Traditional LLMs
Hallucination-Free	✅	❌
Domain-Aware (Legal)	✅	❌
Custom Data Injection	✅	❌
Modular Architecture	✅	⚠️
Streamlit UI	✅	❌

---

🛠️ Installation

git clone https://github.com/yuvraj-kumar-dev/LawgicAI-Chatbot.git
cd LawgicAI-Chatbot
pip install -r requirements.txt

---

▶️ Running the App

1. Create Vectorstore

python create_memory.py

2. Start Chatbot

streamlit run chatbot.py

---

📂 Project Structure

├── data/                     ← Your PDFs go here
├── vectorstore/db_faiss/    ← Vector DB created by FAISS
├── create_memory.py         ← Create memory (RAG setup)
├── connect_memory__.py      ← Connect memory to LLM
├── chatbot.py               ← Streamlit interface
├── LICENSE
└── README.md

---

📃 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

✨ Contribution

If you find this useful, give it a ⭐ and feel free to contribute!

---

👨‍💻 Author

Built with ❤️ by Yuvraj Kumar

---