https://github.com/eitanbakirov/hmo-chatbot
Bilingual chatbot system for Israeli Health Maintenance Organizations (HMOs), using Azure OpenAI and RAG to provide personalized service answers with FastAPI, Streamlit, and vector-based retrieval.
https://github.com/eitanbakirov/hmo-chatbot
azure docker fastapi fullstack-development llm logging monitoring openai python rag streamlit
Last synced: 3 months ago
JSON representation
Bilingual chatbot system for Israeli Health Maintenance Organizations (HMOs), using Azure OpenAI and RAG to provide personalized service answers with FastAPI, Streamlit, and vector-based retrieval.
- Host: GitHub
- URL: https://github.com/eitanbakirov/hmo-chatbot
- Owner: EitanBakirov
- Created: 2025-04-29T12:06:12.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-06-11T12:21:03.000Z (about 1 year ago)
- Last Synced: 2025-10-14T22:39:34.244Z (9 months ago)
- Topics: azure, docker, fastapi, fullstack-development, llm, logging, monitoring, openai, python, rag, streamlit
- Language: HTML
- Homepage:
- Size: 40 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 8
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# HMO Chatbot System
A microservice-based chatbot system for Israeli HMO (Health Fund) service inquiries, using RAG (Retrieval Augmented Generation) and LLM for providing accurate, personalized responses about health services.
The system uses:
- Azure OpenAI (GPT-4o) for chat interactions
- FastAPI for backend services
- Streamlit for frontend interface
- Vector-based document retrieval
- Docker and Docker Compose for deployment
- Enhanced logging and monitoring
## Application Screenshots
### Phase 1
### Phase 2
## Features
- Bilingual support (Hebrew/English)
- Two-phase interaction:
1. User details collection
2. Service inquiries
- RAG-powered answers from HMO documents
- Real-time performance monitoring
- Comprehensive logging system
- Language-aware response formatting
- Docker containerization
## Tech Stack
- Python 3.11
- Azure OpenAI (GPT-4o)
- FastAPI + Uvicorn
- Streamlit
- Docker and Docker Compose
- Vector embeddings for RAG
- Enhanced logging with metrics tracking
## Getting Started
### 1. Clone the repository
```bash
git clone https://github.com/EitanBakirov/health-maintenance-chatbot.git
cd health-maintenance-chatbot
```
### 2. Configure environment
Copy `.env.example` to `.env` and add your Azure OpenAI credentials:
```env
AZURE_OPENAI_KEY=your_key_here
AZURE_OPENAI_ENDPOINT=your_endpoint_here
```
## Project Structure
```
app/
├── backend/ # FastAPI service
│ ├── main.py # API endpoints
│ ├── openai_utils.py
│ ├── retriever.py # RAG implementation
│ ├── initialize.py
│ └── prompts/ # System prompts
├── frontend/ # Streamlit interface
│ └── app.py
├── shared/ # Common utilities
│ ├── logger_config.py
│ └── monitoring.py
├── scripts/ # Data processing
│ └── embed_documents.py
├── phase2_data/ # Source documents
└── docker-compose.yml
```
## Running Options
### Docker Compose
1. Build and run containers:
```bash
docker compose up --build
```
Access:
- Frontend: http://localhost:8501
- Backend: http://localhost:8000
## API Endpoints
- `POST /ask` - Main chat endpoint
- `GET /metrics` - System monitoring
- `GET /health` - Service health check
## Monitoring & Metrics
The system tracks:
- LLM Operations
- Success/failure rates
- Response times
- Token usage
- RAG Performance
- Query similarity scores
- Document retrieval stats
- No-match rates
- Conversation Flow
- Language statistics
- Phase completion rates
- User interaction patterns
## Environment Variables
Required in `.env`:
```env
AZURE_OPENAI_KEY=your_key
AZURE_OPENAI_ENDPOINT=your_endpoint
```
## Example Interaction
1. Language Selection
2. User Information Collection
3. Service Inquiries with RAG
4. Personalized Responses
## Notes
- Requires Azure OpenAI access
- Supports both RTL and LTR text
- Auto-generates embeddings on first run
- Comprehensive logging system
## Potential Upgrades
- Utilize a dedicated VectorDB to save the files
- Add analytics dashboard for monitoring
- Support additional document formats (PDF, DOC)