🤖 AI Study Chatbot with RAG & Exam Generation

A powerful AI-powered study assistant that processes your academic PDFs and enables natural conversation about the content using OpenAI's API and RAG (Retrieval Augmented Generation) technology.

✨ Key Features

• 🤖 Real AI Conversations: Powered by OpenAI GPT-3.5-turbo for natural language understanding
• 📚 Smart PDF Processing: Upload and process academic modules, textbooks, research papers
• 🔍 RAG Implementation: Retrieval Augmented Generation for accurate, document-based responses
• 📝 Intelligent Exam Generation: Create practice tests with 4 difficulty levels from your materials
• 🎯 Document-Specific Responses: AI responses based strictly on your uploaded content
• 💬 Natural Language Interface: Ask questions in plain English about your documents
• 🎓 Educational Focus: Designed specifically for academic study and learning

🚀 Quick Demo

1. Upload your study materials (PDFs)
2. Ask questions like:
   • "What are the main concepts in this document?"
   • "Explain the key principles of [topic] from my uploaded files"
   • "What does Chapter 3 say about [concept]?"
3. Generate exams with customizable difficulty:
   • Easy: Basic recall and definitions
   • Medium: Application and understanding
   • Hard: Analysis and synthesis
   • Expert: Critical thinking and mastery

🛠️ Technology Stack

• AI Engine: OpenAI API (GPT-3.5-turbo + text-embedding-ada-002)
• RAG System: ChromaDB for vector storage with semantic search
• PDF Processing: PyMuPDF, pdfplumber, PyPDF2 with intelligent text extraction
• Backend: Python with modular architecture
• Frontend: Streamlit for intuitive web interface
• Vector Search: OpenAI embeddings with similarity search

📋 Prerequisites

• Python 3.8+
• OpenAI API key (Get one here)
• 2GB+ RAM recommended for vector processing

⚡ Quick Start

1. Clone & Install

git clone https://github.com/yourusername/Study-Chatbot.git
cd Study-Chatbot
pip install -r requirements.txt

2. Configure API Key

# Create .env file
echo "OPENAI_API_KEY=your_openai_api_key_here" > .env

3. Run the Application

streamlit run src/app.py

4. Start Learning! 🎓

• Open http://localhost:8501 in your browser
• Upload your study PDFs
• Start asking questions!

📁 Project Structure

Study-Chatbot/
├── src/
│   ├── app.py              # Streamlit web interface
│   ├── chatbot.py          # Main chatbot orchestration
│   ├── rag_system.py       # RAG implementation with ChromaDB
│   ├── pdf_processor.py    # Advanced PDF text extraction
│   ├── exam_generator.py   # AI-powered exam generation
│   └── config.py           # Configuration management
├── documents/              # Sample documents (optional)
├── requirements.txt        # Python dependencies
├── .env.example           # Environment template
└── README.md              # You are here!

💡 Usage Examples

Chat with Your Documents

You: "What is this document about?"
AI: "This document is a comprehensive guide to Non-Destructive Testing (NDT) 
     methods, covering ultrasonic testing, radiographic inspection, and 
     magnetic particle testing techniques..."

You: "Generate 5 questions about NDT methods"
AI: Creates targeted multiple-choice, true/false, and essay questions
    based on your specific document content.

Advanced Features

• Document Overview: "Summarize the key topics in my uploaded files"
• Specific Queries: "What does section 4.2 say about ultrasonic testing?"
• Comparative Analysis: "Compare the advantages of different NDT methods"
• Exam Generation: Create custom practice tests with answer keys

🔧 Configuration Options

Environment Variables (.env)

# Required
OPENAI_API_KEY=your_openai_api_key_here

# Optional Customizations
OPENAI_MODEL=gpt-3.5-turbo
OPENAI_EMBEDDING_MODEL=text-embedding-ada-002
CHROMA_PERSIST_DIRECTORY=./embeddings
MAX_FILE_SIZE_MB=50

Exam Generation Settings

• Question Types: Multiple choice, True/False, Short answer, Essay
• Difficulty Levels: Easy, Medium, Hard, Expert
• Customizable Counts: Configure questions per type
• Answer Keys: Toggle show/hide functionality

🎯 Core AI Capabilities

RAG (Retrieval Augmented Generation)

1. Document Ingestion: Processes PDFs with advanced text extraction
2. Semantic Chunking: Intelligent text segmentation for optimal retrieval
3. Vector Embedding: OpenAI embeddings for semantic similarity
4. Contextual Retrieval: Finds most relevant document sections
5. Response Generation: AI responses grounded in your content

Intelligent Content Processing

• Multi-format PDF Support: Handles various PDF types and layouts
• Content Quality Filtering: Removes headers, footers, and noise
• Subject-Specific Queries: Optimizes retrieval for technical content
• Overview Generation: Synthesizes document summaries

🚨 Known Limitations

• PDF-only Support: Currently limited to PDF documents
• English Language: Optimized for English-language content
• OpenAI Dependency: Requires active OpenAI API subscription
• Single Session: No persistent user accounts (yet)

🤝 Contributing

We welcome contributions! Areas for improvement:

• Support for more document formats (DOCX, TXT)
• Multi-language support
• User authentication and session persistence
• Advanced analytics and usage tracking
• Collaborative study features

📄 License

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

🙏 Acknowledgments

• OpenAI for providing the GPT and embedding APIs
• ChromaDB for the vector database solution
• Streamlit for the amazing web framework
• LangChain for RAG implementation patterns

📞 Support

• Issues: GitHub Issues
• Documentation: Check the wiki for advanced usage
• Discussions: Share your use cases and get help

---

⭐ Star this repository if it helps with your studies!

Built with ❤️ for students and educators worldwide