Soru.ai is an intelligent educational platform that leverages state-of-the-art AI models to revolutionize how students learn. Upload your handwritten notes or paste a YouTube lecture URL, and watch as our AI generates personalized, high-quality flashcards tailored to your learning style.

Quick Flow:

1. 📤 Upload your handwritten notes or paste a YouTube lecture URL
2. 🔍 AI Processing extracts text via OCR (TrOCR) or transcript (YouTube API)
3. 🧠 Smart Generation creates flashcards using Google Gemini Pro
4. 🃏 Interactive Learning study with clickable, flippable flashcards

Students spend countless hours manually creating flashcards from their notes and lectures. This process is:

• ⏰ Time-consuming: Hours spent on card creation instead of actual studying
• 📝 Inconsistent: Quality varies based on note-taking skills
• 🔄 Redundant: Same content reformatted repeatedly across different courses

Soru.ai automates flashcard generation using:

• OCR Technology: Microsoft's TrOCR model for handwritten text recognition
• AI Generation: Google Gemini Pro for intelligent question-answer pair creation
• Multi-source Support: Process both images and YouTube video transcripts
• Interactive UI: Beautiful, responsive interface built with Reflex

• Upload images of your handwritten notes
• Advanced OCR powered by Microsoft's TrOCR model
• Supports various handwriting styles and formats
• Automatic text extraction and parsing

• Paste any YouTube lecture URL
• Automatic transcript extraction (supports English and Russian)
• AI-powered summarization and concept extraction
• Perfect for MOOC courses and online lectures

• Intelligent question-answer pair creation
• Context-aware content analysis
• Optimized for active recall and spaced repetition
• Character limits for optimal learning (100 chars per side)

• Beautiful gradient design with smooth animations
• Drag-and-drop file upload
• Interactive card flipping
• Real-time processing indicators
• Responsive design for all devices

• Environment-based configuration
• API key protection
• Secure file handling
• No data retention beyond session

┌─────────────────┐
│   User Input    │
│  (Image/URL)    │
└────────┬────────┘
         │
         ▼
┌─────────────────────────────────────┐
│        Soru.ai Platform             │
│  ┌──────────────────────────────┐   │
│  │     OCR Processing           │   │
│  │  (TrOCR Handwritten Model)   │   │
│  └──────────┬───────────────────┘   │
│             │                        │
│             ▼                        │
│  ┌──────────────────────────────┐   │
│  │   YouTube Transcript API     │   │
│  │   (Multi-language Support)   │   │
│  └──────────┬───────────────────┘   │
│             │                        │
│             ▼                        │
│  ┌──────────────────────────────┐   │
│  │    Google Gemini Pro         │   │
│  │  (Flashcard Generation)      │   │
│  └──────────┬───────────────────┘   │
│             │                        │
└─────────────┼────────────────────────┘
              │
              ▼
     ┌────────────────┐
     │  Interactive   │
     │  Flashcards    │
     └────────────────┘

• Python 3.11 or higher
• Google Gemini API key (Get one here)
• Hugging Face API token (Get one here)

1. Clone the repository

   git clone https://github.com/yourusername/soru.ai.git
   cd soru.ai

2. Create and activate virtual environment

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

3. Install dependencies

   pip install -r requirements.txt

4. Configure environment variables

   cp .env.example .env

   Edit .env and add your API keys:

   GOOGLE_API_KEY=your_google_gemini_api_key_here
   HUGGINGFACE_API_KEY=your_huggingface_api_key_here

5. Initialize Reflex

   reflex init

6. Run the application

   reflex run

7. Open your browser

   Navigate to http://localhost:3000
   

soru.ai/
├── lahacks_24/
│   ├── __init__.py
│   └── lahacks_24.py          # Main application logic
├── assets/                     # Static assets (images, icons)
├── uploaded_files/            # Temporary file storage (gitignored)
├── .env                       # Environment variables (gitignored)
├── .env.example              # Environment template
├── .gitignore                # Git ignore rules
├── requirements.txt          # Python dependencies
├── rxconfig.py              # Reflex configuration
└── README.md                # This file

1. Click on the upload area or drag-and-drop your image
2. Select an image of your handwritten notes (JPG, PNG)
3. Click "Upload" button
4. Wait for AI processing (usually 10-30 seconds)
5. View and interact with your generated flashcards

1. Copy a YouTube video URL (e.g., https://www.youtube.com/watch?v=VIDEO_ID)
2. Paste it into the YouTube link input field
3. Click "Generate YouTube Notes"
4. AI extracts transcript and creates flashcards
5. Study your personalized flashcards

class State(rx.State):
    img: list                    # Uploaded image filenames
    cards_list: list[tuple]      # (front, back) flashcard pairs
    visual_cards_list: list      # Currently visible card sides
    processing: bool             # Processing status indicator
    complete: bool               # Completion status

• create_flashcard_prompt(): Processes uploaded images via OCR and generates flashcards
• create_youtube_prompt(): Extracts YouTube transcripts and generates flashcards
• swap_card(): Flips flashcard between front and back
• split_flashcards_list(): Parses AI output into structured flashcard pairs

Beautiful gradient interface with dual input options

Real-time processing with elegant loading indicators

Click any card to flip between question and answer

reflex run --env dev

The project follows Python best practices:

• Type hints for better code clarity
• Async/await for efficient I/O operations
• Clean separation of concerns
• Environment-based configuration
• Secure credential management

We welcome contributions! Please see our Contributing Guide for details.

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

• 🌐 Multi-language support
• 📱 Mobile app development
• 🧠 Additional AI model integrations
• 🎯 Spaced repetition algorithm
• 📊 Learning analytics dashboard
• 🎨 UI/UX improvements

• Handwritten note OCR processing
• YouTube video transcript integration
• Google Gemini flashcard generation
• Interactive card flipping interface
• User authentication and profiles
• Flashcard deck management
• Spaced repetition scheduling
• Export to Anki/Quizlet
• Mobile app (iOS/Android)
• Collaborative study groups
• Performance analytics
• Multi-language UI support
• PDF document processing
• Audio lecture transcription

• No API keys in code: All credentials managed via environment variables
• Secure file handling: Temporary uploads cleaned automatically
• No data persistence: User content not stored on servers
• HTTPS enforced: Secure communication in production

To report security vulnerabilities, please email: [email protected]

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

Soru.ai was created at LA Hacks 2024 by passionate developers who believe in making education more accessible and efficient.

• Project Lead: Your Name
• Contributors: View all contributors

• Reflex - Amazing Python web framework
• Google Gemini - Powerful AI language model
• Microsoft TrOCR - Handwriting recognition
• YouTube Transcript API - Transcript extraction
• LA Hacks 2024 - For the incredible opportunity

• 🌐 Website: soru.ai
• 📧 Email: [email protected]
• 🐦 Twitter: @soruai
• 💬 Discord: Join our community