Comment Crunch is a high-performance analytical tool designed for creators and agencies to "Know Their Community Better." By leveraging Natural Language Processing, it fetches YouTube comment streams and synthesizes them into actionable emotional insights using a modern, editorial interface.
- Real-time Fetching: Seamless integration with the YouTube Data v3 API.
- Neural Sentiment Analysis: Powered by HuggingFace models (RoBERTa) to detect nuanced emotional tones.
- Data Synthesis: Visualizes community feedback through interactive charts and summary metrics.
- Spam Filtering: Automatically identifies and filters rejected comment patterns.
- Modern UX: Minimalist editorial design with fluid Framer Motion transitions and responsive layouts.
- Framework: React + TypeScript (Vite)
- State Management: Zustand
- Styling: TailwindCSS
- Animations: Framer Motion
- Language: Python
- Framework: FastAPI
- Database: PostgreSQL (via Supabase)
- ORM: SQLAlchemy
- AI Engine: HuggingFace Inference API (Sentiment Analysis)
git clone https://github.com/your-username/comment-crunch.git
cd comment-crunch
2. Setup Backend
cd serverpython -m venv venv
venv\Scripts\activate
pip install -r requirements.txt uvicorn main:app --reload
cd frontend
npm install
npm run dev
Environment Variables
Backend (server/.env)
Key Description
DATABASE_URL Your PostgreSQL/Supabase connection string
SECRET_KEY Random string for JWT/Security
YOUTUBE_API_KEY Google Cloud Console API Key
HUGGINGFACE_API_TOKEN HuggingFace Inference API Token
FRONTEND_URL e.g., http://localhost:5173
ENVIRONMENT development or production
Frontend (frontend/.env)
Key Description
VITE_API_URL http://localhost:8000
