This project is a robust, production-ready chatbot system with:

• FastAPI backend for PDF upload, chunking, embedding (Google Gemini 2.0 Flash), vector search, and chat
• MongoDB for vector storage and retrieval
• React frontend with chat UI and admin panel (Netligent branding)
• Environment variable support for all secrets/config
• CORS enabled for frontend integration

1. Clone this repo and navigate to the project directory.
2. Install dependencies:

   pip install -r requirements.txt

3. Set up a .env file with your MongoDB URI, database name, and Gemini API key:

   MONGO_URI=your_mongodb_uri DB_NAME=chatbot_db GEMINI_API_KEY=your_gemini_api_key

4. Run the FastAPI server:

   uvicorn main:app --reload

• POST /upload_pdf — Upload a PDF to update chatbot knowledge
• POST /chat — Ask a question about the course content
• GET /pdfs — List all uploaded PDFs (admin)
• GET /pdf/{id} — View PDF details and chunks (admin)
• DELETE /pdf/{id} — Delete a PDF and its data (admin)

1. Navigate to the frontend directory:

   cd frontend

2. Install dependencies:

   npm install

3. Set the backend API base URL in frontend/.env:

   REACT_APP_API_BASE_URL=http://localhost:8000

4. Start the React app:

   npm start

• Chat UI with Netligent branding and logo
• Admin panel: upload, view, and delete PDFs; view PDF chunks
• Robust error handling and logging
• All secrets/configuration via environment variables

For workspace-specific Copilot instructions, see .github/copilot-instructions.md.