🧠 EduMind Agent

EduMind Agent is an AI-powered autonomous learning and assessment platform that revolutionizes educational content creation and adaptive learning. Built with LangGraph, FastAPI, and Google's Gemini AI, it provides two powerful modes: Teacher Mode for generating comprehensive worksheets and Student Mode for adaptive quizzing and interactive learning.

---

🌟 Features

🎓 Teacher Mode

• Automated Worksheet Generation: Upload PDF study materials (NCERT chapters, reference materials) and automatically generate structured worksheets
• Multi-Format Questions: Creates both Multiple Choice Questions (MCQs) and subjective questions
• Intelligent Topic Segmentation: Automatically identifies and segments topics from uploaded content
• Customizable Output: Configure number of questions, difficulty levels, and question types
• PDF Export: Generates professional PDF worksheets with answer keys
• Smart Question Distribution: Mixed difficulty levels (Easy, Medium, Hard) for comprehensive assessment

🎯 Student Mode

• Adaptive Quiz System: AI-generated quizzes tailored to uploaded study materials
• Configurable Sessions: Customize number of questions, difficulty level, and time limits
• Real-time Feedback: Instant answer validation with explanations
• Performance Analytics: Track accuracy, weak areas, and learning progress
• AI Tutor Chat: Interactive Q&A with context-aware responses from uploaded PDFs
• Enhanced Learning Tools:
  • Mathematical equation solver
  • Function plotting and visualization
  • Automatic follow-up question generation

🔧 Advanced Tools

• Math Solver: Symbolic equation solving using SymPy
• Plot Generator: Dynamic function visualization with matplotlib
• Table Formatter: Clean data presentation
• Diagram Generator: Flowchart creation for process visualization
• Vector Store: Qdrant-based document retrieval for contextual responses

---

🏗️ Architecture

Technology Stack

Backend

• Framework: FastAPI (Python)
• AI/ML:
  • langgraph - Agentic workflow orchestration
  • langchain & langchain-openai - LLM integration
  • google-generativeai - Gemini 2.0 Flash model
  • sentence-transformers - Text embeddings (all-MiniLM-L6-v2)
  • torch - Deep learning framework
• Vector Database: Qdrant Cloud for document storage and retrieval
• Document Processing:
  • pypdf - PDF text extraction
  • python-docx - Word document support
  • openpyxl - Excel file handling
  • pandas - Data manipulation

Frontend

• HTML5 with Jinja2 templating
• CSS3 with modern aesthetics

System Architecture

┌─────────────────────────────────────────────────────────────┐
│                        Frontend                              │
│  (HTML/CSS/JS - Templates/Static Files)                      │
└───────────────────────┬─────────────────────────────────────┘
                        │ HTTP/REST API
┌───────────────────────▼─────────────────────────────────────┐
│                     FastAPI Server                           │
│                      (main.py)                               │
├──────────────────────┬──────────────────────┬───────────────┤
│   Teacher Endpoints  │  Student Endpoints   │ Chat Endpoint │
└──────────────────────┴──────────────────────┴───────────────┘
         │                        │                    │
         ▼                        ▼                    ▼
┌────────────────┐    ┌────────────────────┐   ┌──────────────┐
│ Teacher Agent  │    │  Student Agent     │   │ Chat System  │
│ (LangGraph)    │    │  (LangGraph)       │   │ (RAG-based)  │
└────────┬───────┘    └─────────┬──────────┘   └──────┬───────┘
         │                      │                      │
         ├──────────────────────┴──────────────────────┤
         │                                             │
         ▼                                             ▼
┌─────────────────┐                          ┌─────────────────┐
│  Gemini 2.0     │                          │  Qdrant Vector  │
│  Flash Model    │                          │  Store (Cloud)  │
│  (LLM)          │                          │  (Embeddings)   │
└─────────────────┘                          └─────────────────┘
         │
         ▼
┌─────────────────────────────────────────────────────────────┐
│               Advanced Tools Layer                           │
│  • Math Solver    • Plot Generator    • PDF Exporter        │
│  • Table Formatter    • Diagram Generator                    │
└─────────────────────────────────────────────────────────────┘

Agent Workflows

Teacher Agent Pipeline

PDF Upload → Extract Text → Segment Topics → Generate MCQs → 
Generate Subjective Questions → Format Worksheet → Export PDF

Student Agent Pipeline

PDF Upload → Extract & Embed → Generate Quiz Questions → 
Present Questions → Validate Answers → Analyze Performance

---

📁 Project Structure

Demos/
├── backend/
│   ├── __init__.py
│   ├── config.py              # Configuration & API keys
│   ├── llm.py                 # Gemini LLM interface
│   ├── state.py               # LangGraph state definitions
│   ├── teacher_agent.py       # Teacher mode workflow
│   ├── student_agent.py       # Student mode workflow
│   ├── tools.py               # Advanced tools (math, plots, etc.)
│   └── vector_store.py        # Qdrant vector database interface
│
├── static/
│   ├── css/
│   │   ├── style.css          # Main application styles
│   │   └── chat.css           # Chat interface styles
│   ├── js/
│   │   └── main.js            # Frontend JavaScript
│   └── generated/             # Generated worksheets & outputs
│
├── templates/
│   └── index.html             # Main application template
│
├── temp/                      # Uploaded PDF storage
│
├── main.py                    # FastAPI application entry point
├── requirements.txt           # Python dependencies
├── .env                       # Environment variables (API keys)
└── README.md                  # This file

---

🚀 Getting Started

Prerequisites

• Python 3.8+
• pip (Python package manager)
• Git (for cloning)

Installation

1. Clone the repository

   git clone <your-repository-url>
   cd Demos

2. Create a virtual environment (recommended)

   # Windows
   python -m venv venv
   venv\Scripts\activate
   
   # macOS/Linux
   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies

   pip install -r requirements.txt

4. Configure environment variables

   Create a .env file in the project root:

   GEMINI_API_KEY=your_gemini_api_key_here
   QDRANT_URL=your_qdrant_url
   QDRANT_API_KEY=your_qdrant_api_key

   How to get API keys:

   • Gemini API Key: Visit Google AI Studio
   • Qdrant: Sign up at Qdrant Cloud for free tier

5. Run the application

   python main.py

   The server will start at http://127.0.0.1:8000

6. Access the application

   Open your browser and navigate to http://127.0.0.1:8000

---

📖 Usage Guide

Teacher Mode

1. Upload Study Material

   • Click on "Teacher Mode" in the sidebar
   • Drag & drop a PDF file or click to browse
   • Supported: NCERT chapters, textbooks, reference materials

2. Configure Worksheet Settings

   • MCQ Count: Choose 10, 20, or 50 questions
   • Difficulty: Select Mixed, Easy, or Hard
   • Options:
     • ✅ Include Subjective Questions
     • ✅ Generate Answer Key

3. Generate Worksheet

   • Click "Generate Worksheet"
   • Wait for the AI to process (may take 1-2 minutes)
   • Download the generated PDF worksheet and answer key

4. Output Files

   • Worksheet PDF with all questions
   • Answer key with explanations
   • Markdown version in static/generated/

Student Mode

1. Upload Study Material

   • Click on "Student Mode" in the sidebar
   • Upload your chapter PDF

2. Configure Quiz Settings

   • Number of Questions: 5, 10, 15, or 20
   • Difficulty Level: Easy, Medium, or Hard
   • Time Limit: 5-30 minutes

3. Take the Quiz

   • Click "Start Adaptive Quiz"
   • Answer questions one by one
   • Get instant feedback on each answer
   • View explanations for correct answers

4. Review Performance

   • See your accuracy score
   • Identify weak areas
   • Review missed questions

5. Use AI Tutor Chat

   • After uploading a PDF, the chat interface becomes available
   • Ask questions about the content
   • Get contextual answers with follow-up suggestions
   • Request math solutions, plots, or explanations

AI Chat Features

Example Queries:

• "Explain the concept of photosynthesis"
• "Solve the equation: x^2 - 5x + 6 = 0"
• "Plot the function: sin(x) + cos(x)"
• "Compare mitochondria and chloroplasts in a table"
• "Show me the steps for cellular respiration"

---

🔧 API Endpoints

Public Endpoints

Endpoint	Method	Description
/	GET	Main application interface
/api/upload	POST	Upload PDF file
/api/generate-worksheet	POST	Generate worksheet (Teacher Mode)
/api/student/start	POST	Start quiz session (Student Mode)
/api/student/submit-answer	POST	Submit answer and get validation
/api/student/finish-quiz	POST	Complete quiz and get results
/api/chat	POST	Chat with AI tutor

Request Examples

Upload File:

curl -X POST "http://127.0.0.1:8000/api/upload" \
  -F "file=@chapter.pdf"

Generate Worksheet:

curl -X POST "http://127.0.0.1:8000/api/generate-worksheet" \
  -H "Content-Type: application/json" \
  -d '{
    "filename": "chapter.pdf",
    "mcq_count": 10
  }'

Chat Query:

curl -X POST "http://127.0.0.1:8000/api/chat" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Explain Newton's laws",
    "filename": "physics.pdf"
  }'

---

🧪 Advanced Features

Mathematical Problem Solving

The system can solve equations and evaluate expressions:

# Automatically triggered when queries contain math keywords
Query: "Solve x^2 - 4 = 0"
Response: Solutions: [2, -2]

Function Plotting

Visualize mathematical functions:

Query: "Plot x^2 + 2*x + 1"
# Generates a graph with proper axes and grid

Intelligent Follow-ups

After each answer, the AI generates 3 follow-up questions to deepen understanding:

Answer: [Your Answer]

Follow-up Questions:
1. How does this concept relate to...?
2. Can you explain the practical application of...?
3. What would happen if...?

---

🔐 Security & Configuration

Environment Variables

Variable	Description	Required
GEMINI_API_KEY	Google Gemini API key	✅ Yes
QDRANT_URL	Qdrant cloud instance URL	✅ Yes
QDRANT_API_KEY	Qdrant authentication key	✅ Yes

Security Best Practices

1. Never commit .env file to version control
2. Use environment-specific configurations for development/production
3. Rotate API keys regularly
4. Implement rate limiting for production deployments
5. Sanitize file uploads to prevent malicious files

---

🎨 Customization

Changing the AI Model

Edit backend/config.py:

GEMINI_MODEL_NAME = "gemini-2.0-flash"  # Change to desired model

Adjusting Question Generation

Modify prompts in:

• backend/teacher_agent.py - Worksheet generation
• backend/student_agent.py - Quiz generation

Customizing UI Theme

Edit static/css/style.css to change:

• Color scheme
• Typography
• Layout and spacing
• Animations

---

🐛 Troubleshooting

Common Issues

Issue: "Error reading PDF content"

• Solution: Ensure the PDF is not password-protected and contains extractable text

Issue: "Connection to Qdrant failed"

• Solution: Verify QDRANT_URL and QDRANT_API_KEY in .env file

Issue: "Gemini API rate limit exceeded"

• Solution: Wait a few minutes or upgrade your API quota

Issue: "Torch/CUDA errors"

• Solution: CPU-only mode is sufficient. Install with: pip install torch --index-url https://download.pytorch.org/whl/cpu

Issue: "Module not found"

• Solution: Reinstall dependencies: pip install -r requirements.txt

---

📊 Performance Optimization

For Large PDFs

• PDFs are chunked into smaller segments for better processing
• Vector embeddings are cached in Qdrant
• Consider splitting very large documents (>100 pages)

For Faster Response Times

• Use a local Qdrant instance instead of cloud
• Enable GPU acceleration for torch if available
• Increase API rate limits

---

🤝 Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Development Guidelines

• Follow PEP 8 style guide for Python code
• Add docstrings to all functions
• Test thoroughly before submitting
• Update documentation for new features

---

📝 License

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

---

🙏 Acknowledgments

• Google Gemini AI for powerful language model capabilities
• LangChain & LangGraph for agentic workflow framework
• Qdrant for vector database infrastructure
• FastAPI for modern web framework
• Open source community for various tools and libraries

---