MediBot

A local RAG-based medical chatbot answering questions about cardiovascular and respiratory diseases.

Getting Started

These instructions will get you a copy of MediBot running on your local machine for development and testing purposes.

Prerequisites

The things you need before installing the software.

• Python 3.12+
• Git
• Ollama
• (Recommended) Virtual environment tool (venv or conda)

Installation

1. Clone the repository:

git clone https://github.com/vohuutridung/MediBot.git
cd MediBot

2. Create & activate a virtual environment (optionally)

python -m venv .venv
source .venv/bin/activate # for macOS / Linux

3. Install dependencies

pip install -r requirements.txt

Model Setup

Before running the app, you must have Ollama installed and a local GGUF model available to Ollama.

Environment Variables

Before running evaluation scripts, create a .env file in the project root with the following:

GOOGLE_API_KEY=your_google_api_key_here

This key is required for evaluation tasks that rely on Google APIs.

Usage

1. Start Ollama

ollama serve

2. Go to the project and activate your venv

cd MediBot
source .venv/bin/activate (in macOS / Linux)

3. Run the Streamlit UI

streamlit run src/app.py

Additional Documentation and Acknowledgments

• This is a capstone project for Samsung Innovation Campus (SIC) 2025, developed by a four-member team.
• Read this guide to see how to import models from Hugging Face into Ollama for local usage.