Safar Travel AI Customer Service Agent

An intelligent AI-powered customer service chatbot for Safar Travel, specializing in Iranian domestic travel services.

Features

• Multilingual Support: Communicates in both English and Persian (فارسی)
• Ticket Management: Book, cancel, and retrieve ticket information
• Travel Assistance: Get travel suggestions and recommendations for Iranian destinations
• Company Policies: Access to refund policies, baggage allowances, and booking procedures
• Real-time Search: Web search capabilities for up-to-date travel information
• Knowledge Base: RAG system with company-specific information

Technologies Used

• Frontend: Chainlit for interactive chat interface
• AI Model: OpenAI GPT-4o-mini (configurable)
• Knowledge Base: ChromaDB with sentence transformers
• Search: Tavily API for real-time web search
• Language Processing: Multilingual sentence transformers for Persian/English

Screenshots

Main Chat Interface

Settings Panel

Setup

1. Clone the repository

   git clone <repository-url>
   cd Safar

2. Create virtual environment

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

   pip install -r requirements.txt

4. Configure environment variables Create a .env file in the project root:

   OPENAI_API_KEY=your_openai_api_key_here
   OPENAI_MODEL=gpt-4o-mini
   OPENAI_BASE_URL=https://api.openai.com/v1
   TAVILY_API_KEY=your_tavily_api_key_here

5. Run the application

   python run.py

6. Access the application Open your browser and go to http://localhost:8000

Usage

The AI agent can help you with:

• Booking tickets: "Book a flight from Tehran to Shiraz for tomorrow"
• Cancelling tickets: "Cancel ticket ST-1001"
• Getting ticket info: "Show me details for ticket ST-1001"
• Travel suggestions: "Suggest places to visit in Isfahan"
• Company policies: "What is your refund policy?"

Both English and Persian queries are supported.

Project Structure

Safar/
├── chainlit_app.py      # Main Chainlit application
├── run.py               # Application launcher
├── src/
│   ├── manager.py       # Agent orchestration and LLM calls
│   └── tools.py         # Tool implementations (RAG, API, Search)
├── data/               # Knowledge base documents
├── prompts.py          # System prompts and few-shot examples
├── requirements.txt    # Python dependencies
└── .env               # Environment configuration

License

This project is for educational and demonstration purposes.