The DocBot: AI-Powered Medical Assistant

The DocBot is an AI-powered medical assistant designed to provide accurate and contextually relevant medical assistance in response to user queries. This project integrates advanced natural language processing (NLP) techniques with a user-friendly frontend to deliver a seamless experience.

Features

1. Advanced AI Chatbot

Built using Mistral-Nemo-Instruct-2407, deployed through the HuggingFace Inference API.
Offers precise and informed responses to medical-related queries.

2. Comprehensive Medical Knowledge Base

Integrated 5 volumes of medical encyclopedias with over 2,500 pages of content.
Utilizes FAISS vector database to enable Retrieval-Augmented Generation (RAG) for high-quality answers.

3. User-Friendly Interface

ReactJS frontend for an intuitive user experience.
Google Authentication for secure user access and profile management.

4. Personalized Responses

Stores user profiles in a database to enhance chatbot response quality by tailoring answers to user-specific needs.

Technologies Used

AI Model: Mistral-Nemo-Instruct-2407 via HuggingFace Inference API
Database: FAISS (Facebook AI Similarity Search)
Frontend: ReactJS
Authentication: Google Authentication
Backend: FastAPI, SQLite db for user profile storage

How to Use

Clone the repository.

git clone https://github.com/Udit-Krishna/The-DocBot.git

Install dependencies for the Backend

cd Backend
python -m venv <venv name>
.\<venv name>\Scripts\activate
pip install -r requirements.txt

Start the backend server.

uvicorn app:app

Install dependencies for the frontend.
```
 cd Frontend
 npm install
```
Start the backend server.

  uvicorn app:app

Start the frontend.
```
npm run dev
```
Configure the backend with appropriate API keys and database credentials.

Feel free to reach out if you have any questions or suggestions!