A full-stack AI-powered chatbot that can read and answer questions from uploaded PDF documents in multiple languages (English, Hindi, Telugu).
- 🔺 Upload PDF documents via an elegant frontend.
- 🤖 Ask questions and receive answers based on PDF content.
- 🧠 Supports multilingual queries: English, Hindi, Telugu.
- 📝 Generate concise summaries from PDF files.
- 🗑️ Delete uploaded PDFs easily from the interface.
- ⚡ Built with FastAPI, LangChain, React.js, and Ollama for local LLM support.
| Backend | Frontend |
|---|---|
| FastAPI + LangChain | React.js |
| Ollama (Mistral Model) | Axios (API requests) |
| FAISS (Vector DB) | Responsive UI with CSS |
pdf\_chatBot/
├── backend/
│ ├── main.py
│ └── requirements.txt
└── frontend/
└── src/
├── App.jsx
└── App.css
- Python 3.9+
- Node.js & npm
- Ollama installed locally (for Mistral model)
cd backend
python -m venv venv
source venv/Scripts/activate # On Windows
pip install -r requirements.txt
uvicorn main:app --reloadMake sure Ollama is running and has the mistral model downloaded:
ollama run mistralcd frontend
npm install
npm startThe React app will start at http://localhost:3000.
- Upload a
.pdffile. - Select a language (English, Hindi, Telugu).
- Ask questions or click "Summarize" to get the main idea of the PDF.
- Switch between PDFs and delete them anytime.
Vulkunda Srinath 📧 Email: srinathvulkunda@gmail.com 🌐 GitHub: @Srinath-Vulkunda
This project is open-source and free to use. You can modify and extend it for educational or personal purposes.
- Login and authentication
- Support for multiple file types
- Persistent vector DB (e.g., Pinecone or Chroma)
- Improved translation accuracy