Logo Lanfrica

Respectus11/Chatbott

Domain:

healthcarenatural language processing

Record type:

software
Creator:
Res
Host:
ምርኩዜ Merkuze -A hospital AI chatbot designed to assist patients and visitors at Tikur Anbessa Specialized Hospital in Ethiopia. # ምርኩዜ Merkuze - Tikur Anbessa Hospital AI Assistant A hospital AI chatbot powered by Ollama (gemma:3.1b) designed to assist patients and visitors at Tikur Anbessa Specialized Hospital in Ethiopia. The chatbot provides information about departments, doctors, appointments, and other hospital services through an intuitive and modern interface. ## Table of Contents - Overview - Features - Technologies Used - Project Structure - API Endpoints - Getting Started - Configuration - Demo Screenshots - Future Enhancements ## Overview Merkuze (ምርኩዜ) is an intelligent hospital assistant that helps users navigate healthcare services at Tikur Anbessa Specialized Hospital. Using Retrieval-Augmented Generation (RAG) technology, the chatbot provides accurate and context-aware responses about hospital departments, medical staff, appointments, and general health inquiries. ## Features ### AI-Powered Conversations - Powered by Ollama with gemma:3.1b model - Retrieval-Augmented Generation (RAG) for accurate hospital information - Multilingual support (English and Amharic) - Natural and context-aware responses ### User Authentication - Secure signup and login with JWT tokens - Password hashing with bcrypt - Role-based access control (User/Admin) - Session persistence ### Chat Features - Real-time chat interface - Chat history storage and retrieval - Start new conversations - Resume previous chats - Typing indicators for better UX ### Admin Dashboard - Document upload for knowledge base enhancement - Manage uploaded documents - Admin-only access control ### User Experience - Modern and responsive interface - Floating chat widget on landing page - Smooth animations and transitions - Hospital themed color Used ### Backend scheme ## Technologies - **Runtime**: Node.js - **Framework**: Express.js - **Database**: MongoDB - **Authentication**: JWT (JSON Web Tokens) - **Password Hashing**: bcrypt - **File Upload**: Multer - **AI Integration**: Ollama (gemma:3.1b) ### Frontend …

Languages