🕉️ Bhagavad Gita GPT

A conversational AI that answers questions about the Bhagavad Gita using Retrieval-Augmented Generation (RAG). Ask anything about the Gita's teachings, verses, and philosophy — and get accurate, context-grounded answers.

---

✨ Features

• 📖 PDF-based knowledge — answers grounded strictly in the Bhagavad Gita text
• 🧠 Chat memory — remembers previous questions within a session
• 🔍 Source transparency — shows which verses/pages the answer came from
• 🚫 Hallucination guard — only answers from the document, never from general knowledge
• ⚡ Fast responses — powered by Groq's ultra-fast LLM inference
• 💾 Persistent vector store — FAISS index built once, reused every run

---

🛠️ Tech Stack

Layer	Technology
UI	Streamlit
LLM	Groq (llama-3.3-70b-versatile)
Embeddings	HuggingFace (all-MiniLM-L6-v2)
Vector Store	FAISS (local)
RAG Framework	LangChain
PDF Loader	PyPDFDirectoryLoader

---

📦 Installation

1. Clone the repository

git clone https://github.com/yourusername/BhagavadGitaGPT.git
cd BhagavadGitaGPT

2. Create and activate virtual environment

python -m venv myenv

# Windows
myenv\Scripts\activate

# Mac/Linux
source myenv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables

Create a .env file in the root directory:

GROQ_API_KEY=your_groq_api_key_here

Get your free Groq API key at → https://console.groq.com

5. Add your PDFs

Place your Bhagavad Gita PDF(s) inside the pdfs/ folder:

BhagavadGitaGPT/
└── pdfs/
    └── bhagavad_gita.pdf

---

🚀 How to Use

First run — builds the vector store automatically:

streamlit run BhagavadGita.py

The app will:

1. Load PDFs from the pdfs/ folder
2. Chunk and embed the text
3. Save the FAISS index to vector_store/
4. Launch the chat interface at http://localhost:8501

Subsequent runs — loads from disk instantly:

streamlit run BhagavadGita.py

No re-ingestion needed — the vector store is reused automatically.

---

📁 Project Structure

BhagavadGitaGPT/
│
├── pdfs/                   # Place your Bhagavad Gita PDFs here
├── images/                 # UI images (krishna.png, sacred_book_1.png)
├── vector_store/           # Auto-generated FAISS index (gitignored)
├── BhagavadGita.py         # Main Streamlit app
├── requirements.txt        # Python dependencies
├── .env                    # API keys (gitignored)
└── .gitignore

---

⚙️ Requirements

streamlit
langchain==0.3.15
langchain-community==0.3.15
langchain-groq
langchain-huggingface
langchain-text-splitters
langchain-core
sentence-transformers
faiss-cpu
pypdf
python-dotenv

🙏 Acknowledgements

• Bhagavad Gita — the sacred text
• LangChain — RAG framework
• Groq — LLM inference
• Streamlit — UI framework