Experimental Seat Booking Agent

Experimental seat booking agent implementation combines OpenAI Agents SDK, Autogen extensions, and MCP tools support.

Use Case

The self-service airline seat reservation kiosk, also known as a flight seat booking service agent or passenger service agent, is designed to provide comprehensive customer support and assistance. This kiosk ensures a smooth and positive travel experience by handling various tasks, including reservations, through its call center interface.

Key Features:

• Triage agent answer user general inquiry with tools
• Triage agent should route customer question to specialized agents
• Triage agent retrieve user preferences on seating, airlines and food
• Triage agent use agents as tools on question routing
• Triage agent use MCP server as tools to perform web fetch content and read files system
• QA agent has knownledge to handling questions about Plane, Flight and Onboarding
• Seat booking agent specifilized on plane seat reservation tasks
• Web Chat interface for user self-service

Lab Technical Stack

• Linux
• OpenAI Agents SDK
• Autogen 0.4 + extensions (MCP Tools,OpenAI, Ollama)
• Chainlit (UI)
• Ollama (local Qwen2,5)

Environment

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Running

chainlit run app.py

Todos

• save seat history at the end of the booking
• improve FAQ agent to use RAG on actual documents
• improve memory storage
• add MCP tools to invoke actual booking service api
• more to came...

Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

Important: Please keep your pull requests small and focused. This will make it easier to review and merge.

Feature Requests

If you have a feature request, please open an issue and make sure it is tagged with enhancement.

License

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