Nyay-Saarthi is an AI-powered legal assistant designed to simplify complex legal documents for users in both Hindi and English.
This application allows users to upload legal documents and receive simplified summaries, ask questions about the content, and even book consultations with legal experts.
- Document Upload and Simplification: Securely upload legal documents (PDF, DOCX, TXT, PNG, JPG, JPEG) and receive a simplified summary in plain language.
- AI-Powered Q&A: Ask questions about your uploaded documents and get instant, context-aware answers from our AI assistant.
- Consultation Booking: Schedule consultations with legal experts through the platform.
- User Authentication: Secure user registration and login functionality.
- Dashboard: An overview of your documents and analysis.
- *Bilingual Support: The entire user interface and AI responses are available in both *Hindi and English.
- Responsive Design: A mobile-friendly interface ensures a seamless experience across all devices.
- Framework: Next.js
- Language: TypeScript
- Styling: Tailwind CSS with shadcn/ui components
- Authentication: Supabase
- Framework: FastAPI
- Language: Python
- AI/ML: LangChain, Google Generative AI, Hugging Face Embeddings
- Vector Store: Qdrant
- Deployment: Docker
- Node.js and npm (or yarn/pnpm)
- Python 3.11
- Docker
-
Clone the repository: bash git clone https://github.com/parrth20/nyay-saarthi.git cd nyay-saarthi
-
Frontend Setup: bash
npm install
-
Backend Setup: bash cd Backend pip install -r requirements.txt
-
Environment Variables:
Create a .env file in the root of the frontend and backend directories.
Add the necessary environment variables (e.g., Supabase URL and keys, Google Generative AI API key).
Start the Frontend: bash npm run dev
Start the Backend: bash cd Backend uvicorn main:app --host 0.0.0.0 --port 8000
Alternatively, run the backend using Docker: bash cd Backend docker build -t nyay-saarthi-backend . docker run -p 8000:8000 nyay-saarthi-backend
Contributions are welcome!
Please feel free to open an issue or submit a pull request.