This work discusses some approaches to set up an assistant system for the new product IVU.Plan essentials of the German company IVU Traffic Technologies. The assistant should be able to help modify a timetable for public transport by taking instructions in natural language, understanding the intent of the user, and fulfilling it. The most successful of the shown approaches delivers less than 1% rate of critical misinterpretations. Furthermore, the relatively new Message Context Protocol (MCP) is introduced and integrated into a prototype of an assistant. It is used to give a Large Language Model (LLM) the possibility to look up current timetable data on its own from an internal data source. This work will develop a tool which improves this data retrieval by massively reducing the data size and thus the cost of the LLM use without losing any precision. The setup will be able to answer nearly 100% of questions regarding the timetable correctly, which gives the assistant the power to fulfill the intent of the user appropriately. This work lays the foundation for the future development of an assistant for the product IVU.Plan essentials and shows the massive potential of the Message Context Protocol - also for other applications.