PadhAI - Prompt-to-Lecture Generator

An interface for generating educational lectures with synchronized animations and narration using AI.

Features

• Friendly UI: Three-panel layout with sidebar, video player, and chat interface
• AI-Generated Content: Uses Groq API (OSS model) to generate Manim code and narration scripts
• Manim Animations: Automatically generates educational animations for STEM topics
• TTS: Browser Web Speech API (free)
• Video Output: Combines animations with narration for complete lecture videos
• Keyboard Shortcuts: Enter to submit, Shift+Enter for new line

Project Structure

gpt-open-hack/
├── frontend/          # React + Vite frontend
│   ├── src/
│   │   ├── components/
│   │   │   ├── Sidebar.jsx      # Left panel - chat history
│   │   │   ├── VideoPlayer.jsx  # Center panel - video player
│   │   │   └── ChatPanel.jsx    # Right panel - prompt input
│   │   └── App.jsx
│   └── package.json
├── backend/           # Node.js + Express backend
│   ├── routes/
│   │   └── generate.js          # Main API endpoint
│   ├── videos/                  # Generated video files
│   └── index.js
└── README.md

Setup Instructions

Prerequisites

1. Node.js (v20+)
2. Python with Manim installed:

   pip install manim

3. Groq API Key (or adapt back to OpenRouter)

Installation

1. Clone and setup:

   cd gpt-open-hack

2. Setup Backend:

   cd backend
   npm install
   copy .env.example .env   # (Windows PowerShell; use cp on macOS/Linux)
   # Edit .env and add your GROQ_API_KEY

3. Setup Frontend:

   cd ../frontend
   npm install

Configuration

1. Get Groq API Key:

   • Visit https://console.groq.com/
   • Create an API key

2. Configure Backend (backend/.env):

   PORT=3001
   GROQ_API_KEY=your_key_here

Note: TTS is handled automatically by your browser - no additional setup required!

Running the Application

1. Start Backend (Terminal 1):

   cd backend
   npm start

2. Start Frontend (Terminal 2):

   cd frontend
   npm run dev

3. Open Browser:

   • Navigate to http://localhost:5173

Usage

1. Enter a Prompt: In the right panel, enter a topic like:

   • "Explain Newton's laws of motion with visual examples"
   • "Teach calculus derivatives with animated graphs"
   • "Show how electromagnetic waves work"

2. Generate Lecture: Click "Generate Lecture" to create:

   • Manim animation code
   • Narration script
   • Combined video output

3. View Results:

   • Video plays in center panel
   • Chat history appears in left panel
   • Click previous chats to replay videos

Technical Details

Frontend (React + Vite)

• Components: Modular React components for each panel
• Styling: VS Code-inspired dark theme with CSS
• State Management: React hooks for managing video and chat state

Backend (Node.js + Express)

• API Endpoint: /api/generate - handles lecture generation
• LLM Integration: Groq API for content generation (model: openai/gpt-oss-120b)
• Manim Execution: Runs Python scripts to generate animations
• File Serving: Static file serving for generated videos

AI Integration

• Model: openai/gpt-oss-120b via Groq
• Prompt Engineering: "Maestro" system prompt producing shot-based timeline JSON (title, manim_header, scenes[shots], manim_footer)
• Output Format: JSON transformed into executable Manim scene

Free Services Used

• Groq: Fast OSS model inference API
• Browser TTS: Web Speech API for text-to-speech
• Local Manim: Community edition for animations
• Local Storage: Browser memory for chat history

Troubleshooting

Common Issues

1. Manim not found:

   pip install manim
   # or
   conda install -c conda-forge manim

2. API Key errors:

   • Check .env file exists in backend/
   • Verify GROQ_API_KEY is correct
   • If switching back to OpenRouter, adjust endpoint & env var in routes/generate.js

3. Port conflicts:

   • Backend runs on port 3001
   • Frontend runs on port 5173
   • Change ports in config if needed

Development Notes

• Backend includes fallback responses for offline development
• Minimal guardrails (prompt length + pattern filtering) applied before execution
• For production consider sandboxing Manim (Docker, seccomp, etc.)

Future Enhancements

• Real TTS streaming integration
• Audio + video mux (ffmpeg) for packaged lectures
• Template library of common pedagogy animations
• Persistent storage for prompts & metadata
• Authentication / multi-user support
• Docker sandbox / safer code execution
• CI (lint, type, minimal unit test)

Contributing

1. Fork the repository
2. Create feature branch
3. Make changes
4. Test locally
5. Submit pull request

License

MIT License - see LICENSE file for details.