BeyondQuiz 🎓

AI-powered quiz generation platform that helps students learn from their coursebooks through interactive quizzes, intelligent chat assistance, and curated video recommendations.

🚀 Live Demo: https://beyond-quiz.vercel.app/

---

Features

Quiz Generation

• Upload PDF coursebooks or select from pre-loaded NCERT textbooks
• AI generates three question types: MCQs, Short Answer, and Long Answer questions
• Intelligent scoring with detailed feedback and explanations
• Progress tracking dashboard with performance analytics

AI Teacher Chat

• ChatGPT-style interface for document-based Q&A
• Context-aware responses based on uploaded materials
• Persistent chat history across sessions
• Multiple concurrent chat sessions

Video Recommendations

• AI-curated educational YouTube videos based on document content
• Topic-specific recommendations for enhanced learning

Progress Tracking

• Comprehensive dashboard showing quiz attempts and scores
• Performance trends and improvement metrics
• Detailed attempt history with feedback summaries

---

Tech Stack

Backend: Flask, PyMongo, Google Gemini 2.5 Flash, PyPDF2
Frontend: HTML5, CSS3, JavaScript, PDF.js
Database: MongoDB Atlas
Deployment: Vercel

---

Quick Start

Prerequisites

• Python 3.9+
• MongoDB Atlas account
• Google Gemini API key

Installation

1. Clone the repository:

git clone [your-repo-url]
cd beyondquiz

2. Install dependencies:

pip install -r requirements.txt

3. Configure environment variables: Create a .env file:

MONGO_URI="mongodb+srv://username:password@cluster.mongodb.net/dbname"
GEMINI_API_KEY="your_gemini_api_key"

4. Run the application:

python app.py

5. Open http://127.0.0.1:5000 in your browser

---

Deployment

Deploy to Vercel:

vercel

Configure environment variables in Vercel Dashboard:

• MONGO_URI
• GEMINI_API_KEY

---

Project Structure

beyondquiz/
├── app.py                 # Flask backend with API endpoints
├── requirements.txt       # Python dependencies
├── vercel.json           # Vercel configuration
├── templates/
│   └── index.html        # Main application template
└── static/
    ├── style.css         # Application styles
    ├── chat-style.css    # Chat interface styles
    ├── script.js         # Application logic
    └── chat-script.js    # Chat logic

---

API Endpoints

• POST /api/upload - Upload PDF coursebook
• GET /api/pdfs - List all uploaded PDFs
• POST /api/generate-quiz - Generate quiz from PDF
• POST /api/score-quiz - Score quiz submission
• POST /api/chat - Chat with AI teacher
• POST /api/recommend-videos - Get video recommendations
• GET /api/progress - Retrieve quiz history

---

Known Limitations

• PDF visual preview not persistent after refresh (serverless constraint)
• No user authentication (global progress tracking)
• Gemini API rate limits on free tier (handled with retry logic)
• Large PDFs (500+ pages) may experience longer processing times

---

Future Enhancements

• User authentication and personalized progress
• Cloud storage for persistent PDF previews
• Vector embeddings for semantic search
• Export progress reports
• Spaced repetition scheduling
• Mobile application

---

License

Apache License 2.0