PDF Question-Answering Chatbot 🤖

📌 Introduction

The PDF Question-Answering Chatbot is an AI-powered assistant designed to help users extract information from multiple PDF documents. This versatile tool allows you to upload one or more PDFs and ask questions about their content, receiving accurate and contextual responses. The system is built with Streamlit, LangChain, FAISS, and leverages state-of-the-art LLMs to generate dynamic responses.

🔗 Live App: course-faq-bot.streamlit.app

✨ Features

✅ Process and analyze multiple PDF documents in a single session
✅ Ask questions about the combined content of uploaded PDFs and receive AI-generated answers
✅ Uses FAISS for fast and efficient vector search capabilities
✅ Supports Groq and OpenAI models for natural language processing
✅ Clean, interactive UI built with Streamlit
✅ Secure API key management via Streamlit Expander
✅ Document chunking for optimal context retrieval

🛠️ Tech Stack

• Python
• Streamlit (for user interface)
• LangChain (for document processing and retrieval chains)
• FAISS (for vector storage and similarity search)
• OpenAI Embeddings (for text embedding)
• Groq LLM (for response generation)

🚀 How It Works

1. Upload one or more PDF documents through the Streamlit interface.
2. The application splits the documents into manageable chunks.
3. These chunks are embedded using OpenAI's embedding model.
4. FAISS stores these embeddings in a vector database for efficient retrieval.
5. When you ask a question, the system retrieves the most relevant document sections from all uploaded PDFs.
6. Groq's LLM processes these sections along with your question to generate a contextual response.
7. The chatbot presents answers in a conversational format, asking if you need additional information.

🏗️ Project Structure

FAQ_BOT/
│── data/                  # Temporary storage for uploaded PDF files
│── faiss_db/              # Persisted FAISS database
│── app.py                 # Main Streamlit application
│── requirements.txt       # Python dependencies
│── README.md              # Documentation
│── .env                   # Environment variables (API keys)

🏗️ Installation & Setup

🌐 Using the Deployed App

The easiest way to try the app is to visit the live deployment:

• Live App: course-faq-bot.streamlit.app

1️⃣ Clone the Repository (For Local Development)

git clone https://github.com/Shafin008/FAQ_BOT.git
cd FAQ_BOT

2️⃣ Create a Virtual Environment

python -m venv venv
source venv/bin/activate  # On macOS/Linux
venv\Scripts\activate    # On Windows

3️⃣ Install Dependencies

pip install -r requirements.txt

4️⃣ Set Up API Keys

You'll need API keys for both services:

• Groq API Key: Get it here
• OpenAI API Key: Get it here

You can either:

• Add them to a .env file in the project root
• Input them directly in the Streamlit interface when prompted

5️⃣ Run the App

streamlit run app.py

🤖 Using the Chatbot

1. Enter Your API Keys: Provide your Groq and OpenAI API keys in the password fields (click "Click here to enter your API KEYs").
2. Upload PDFs: Use the file uploader to upload one or more PDF documents you want to query.
3. Start Engine: Click the "Start Engine" button to process your documents into the vector database.
4. Ask Questions: Use the chat input to ask specific questions about the content of the uploaded PDFs.
5. Get Answers: Receive contextual responses based on the combined content of all uploaded documents.
6. Follow-up: The bot will ask if you're satisfied or need more information.

💡 Example Use Cases

• Academic Research: Extract specific information from multiple research papers simultaneously.
• Legal Document Analysis: Query multiple legal documents for specific clauses or information.
• Technical Documentation: Ask questions about multiple technical manuals or guides at once.
• Course Materials: Learn from multiple educational PDFs with interactive Q&A.
• Financial Reports: Extract insights from multiple financial documents through natural language queries.

🔮 Future Enhancements

✅ Add support for multiple document formats (DOCX, TXT, etc.).
✅ Implement document summarization capabilities.
✅ Add option to compare information across multiple documents.
✅ Create visualization tools for document insights.
✅ Deploy as a standalone web service with user accounts.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request or open an Issue if you have suggestions or encounter problems.

📜 License

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

📬 Contact

If you have any questions or suggestions, feel free to reach out:

📧 Email: shafinmahmud114@gmail.com
🐦 Twitter: @shafinmahmud114
💼 LinkedIn: Shafin Mahmud Jalal

---

⭐ If you find this project helpful, please consider giving it a star!