🏥 CliniWise

A modern, AI-powered clinical guidelines management system that helps healthcare professionals access and interact with medical documentation efficiently.

🌟 Features

• Template of the application is very heavily inspired from the SEC insights repo (https://github.com/run-llama/sec-insights)

• 🤖 AI-Powered Interactions

  • Natural language querying of medical documents
  • Context-aware responses based on document content
  • Intelligent document summarization

• 🔍 Advanced Search

  • Full-text search across all documents
  • Semantic similarity matching
  • Filter by document type and metadata

• 👥 User-Friendly Interface

  • Modern, responsive web interface
  • PDF viewer with highlighting
  • Intuitive conversation interface

• Examples guidelines uploaded to application for you to get started with

🚀 Getting Started

Prerequisites

• Docker and Docker Compose
• Python 3.10 or later
• Node.js 16 or later
• Poetry (Python package manager)

Installation

1. Clone the repository

   git clone https://github.com/yourusername/cliniwise.git
   cd cliniwise

2. Set up the backend

3. Set up the frontend

   cd frontend
   npm install
   npm run dev

4. Start the services

   # In the backend directory
   docker-compose up -d
   make migrate
   make run
   
   # In a new terminal, frontend directory
   npm run dev

5. Initialize the database

   make seed_db_local

🏗️ Architecture

CliniWise is built with a modern tech stack:

• Frontend: React, TypeScript, Vite
• Backend: FastAPI, PostgreSQL, pgvector
• AI/ML: LlamaIndex for document processing and embeddings
• Storage: S3-compatible storage (LocalStack for development)

📝 Development

Local Development Setup

1. Start the local services:

   make run

2. Seed the database with example clinical guidelines:

   make seed_db_local

3. Access the application:

   • Frontend: http://localhost:3000
   • API: http://localhost:8000
   • API Documentation: http://localhost:8000/docs

Environment Variables

Key environment variables:

DATABASE_URL=postgresql://user:password@localhost:5432/llama_app_db
S3_ENDPOINT_URL=http://localhost:4566
S3_ASSET_BUCKET_NAME=clinical-guidelines-assets
OPENAI_API_KEY=your-api-key

🤝 Contributing

I welcome contributions! Please see our Contributing Guide for details.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with LlamaIndex
• PDF processing powered by PyMuPDF
• Very heavily inspired from the SEC insights repo (https://github.com/run-llama/sec-insights)