What is agent-customerservice?

This project is a hotel customer service chatbot built in Python, with a README written entirely in Chinese. The system is designed to handle the kinds of requests a hotel front desk typically manages: room reservations, information questions, membership account services, weather lookups, and the option to hand off to a human agent when needed. The project name roughly translates to "Zhixiang Hotel Intelligent Customer Service System." The system is organized as a set of specialized AI agents, each responsible for a specific task type. A coordinator agent figures out what the user is asking and routes it to the right specialist: a booking agent, an information-query agent that searches a knowledge base, a member service agent for account and loyalty points, a weather agent, or a general chat agent. Bookings are stored in SQLite, and the system uses Redis to cache repeated queries. The knowledge base for FAQ and facility information is built with ChromaDB, a vector search tool that lets the system find relevant answers by meaning rather than exact keyword match. The AI language model powering the responses is Tongyi Qianwen, Alibaba's model, accessed via API key. The web interface runs on Streamlit, a Python library that creates simple browser-based interfaces. There is also a separate Streamlit page for a document-cleaning tool that converts uploaded PDF, Word, or text files into Markdown format, which the system can then load into its knowledge base. To run the project you need Python 3.8 or newer, a Qianwen API key configured in a YAML file, and then you start the Streamlit app on a local port. The project is released under the MIT license.