An AI-driven PDF RAG Chat Application
The AI RAG PDF App is a cutting-edge AI-powered document interaction tool designed to enhance collaboration, retrieval-augmented generation (RAG), and AI-assisted chat with PDFs. Users can upload PDFs, extract insights using multimodal LLMs, and collaborate in shared spaces.
- Multi-Model Support: Users can chat with PDFs using state-of-the-art models like LLAMA, Mixtral, and others via the Groq API.
- Spaces Collaboration: Generate a unique space from a PDF where multiple users can collaborate in real time, viewing and contributing to the chat history.
- Collection Management: Create a collection of multiple PDFs (like a playlist) and chat with them individually for a structured knowledge retrieval experience.
- Enhanced Contextual AI: The app uses Pinecone as a vector database for efficient document search and retrieval.
- Fast & Efficient Embeddings: Hugging Face embedding inference ensures high-quality text vectorization for better responses.
- Frontend: React.js (Vite)
- Backend: Node.js with Express.js
- Database: MongoDB
- AI Models: Groq API (LLAMA, Mixtral, etc.)
- Vector Database: Pinecone
- Embeddings: Hugging Face inference API
- Authentication: JWT (JSON Web Token)
- Storage: Multer (for PDF uploads)
Ensure you have the following installed:
- Node.js (>= 16.x)
- MongoDB
- Pinecone Account (for vector storage)
- Hugging Face API key
- Groq API key
- Clone the Repository:
git clone https://github.com/your-repo/ai-rag-pdf.git cd ai-rag-pdf # root directory
- Install Dependencies:
cd api npm installcd client npm install - Setup Environment Variables:
Create a
.envfile in the root directory and add the following:PORT=3000 MONGODB_URI=mongodb://localhost:27017/your-ragapp JWT_SECRET="your_secret_key" GROQ_API_KEY="your_groq_api_key" PINECONE_API_KEY="your_pinecone_api_key" PINECONE_INDEX="your_pinecone_index" HUGGINGFACE_API_KEY="your_huggingface_api_key"
- Run the Application: Backend
The server will start on
cd api npm run devhttp://localhost:3000. - Run the Application: Frontend
The frontend will be available at
cd client npm run dev npm run dev -- --host # For mobile access on same Wi-Fi
http://localhost:5173.
| Method | Endpoint | Description |
|---|---|---|
| POST | /signup | Register a new user |
| POST | /login | Authenticate user |
| POST | /upload | Upload a new PDF |
| GET | /documents | Get all PDF |
| GET | /query | Chat with a specific PDF |
| POST | /spaces/create | Create a collaborative space |
| GET | /spaces/join | Access a shared PDF chat space |
| GET | /spaces/ | Get user's spaces |
| GET | /spaces/:id | Get specific space by ID |
| POST | /collection/create | Create a PDF collection |
| GET | /collection/:collectionId | Retrieve a collection |
updating...
- Recommended: Best experienced on desktop for full feature access.
- Collaborative Mode: Ensure all participants have access to the shared space link.