Request ERP information directly in Microsoft Teams

Microsoft intends for ERP data to be accessible directly in Microsoft Teams. This is the result of a new technical foundation in Microsoft Dynamics 365 Business Central, namely the so-called MCP Server.

This server acts as a kind of interpreter between the AI chatbot and the business logic of Business Central. In other words, the server ensures that AI does not access data “somehow,” but works within the ERP system based on rules and understands queries even if no ERP technical vocabulary is used in the chat.

With the MCP Server, Business Central has created a clean foundation for building AI chatbots and AI agents for the first time. Other application scenarios besides warehousing include sales (Sales Order Agent), service, purchasing, and supply chain.

Technically speaking, the agents do not access Business Central tables, but ask their questions via the MCP Server. From there, they receive context-related answers according to Business Central logic. Microsoft Teams serves merely as an interface. Employees can ask questions in this way without opening the ERP system – and still receive valid, up-to-date information directly from the ERP.

Warehouse employees need one thing above all else in their daily work: fast, reliable information. How many items are available? Where is the stock located? Is there enough quantity for the next order? These are precisely the questions that often arise under time pressure – and require a real-time answer without having to spend time “looking it up in the system.”

Teams chatbot for Microsoft Dynamics 365 Business Central brings this information to where warehouse employees already work and communicate: in Microsoft Teams. This saves time, clicks, and a lot of effort.

The advantages of Teams chatbots for warehouse staff at a glance

No need to switch to the ERP system
Instead of logging into Business Central, navigating through masks, or searching for item numbers, a short message in Teams is all it takes. The answer comes directly from the ERP system.

Natural language instead of item numbers
Even warehouse employees don't have all the item numbers in their heads. They also ask for “the chair,” “the red version,” or “the item from last week.” The search function in the ERP would be overwhelmed by this. The AI agent, on the other hand, understands what is meant and translates the queries for the ERP system into system language.

Dialogue instead of a one-time question
If information is missing or there are several possible hits, the agent asks specific questions. This creates a professional dialogue – comparable to an experienced warehouse colleague who asks questions instead of making false assumptions. 

Ideal for mobile situations
Thanks to its very simple interface, Teams is significantly faster to access and easier to use than Business Central on smartphones. The chatbot therefore makes warehouse information more accessible, especially on mobile devices.

Teams chatbot for Microsoft Dynamics 365 Business Central is neither a classic search field nor a general AI assistant with unlimited knowledge. It works as a specialized agent within the logic of Business Central and is deliberately geared toward warehouse processes. This limitation is not a disadvantage, but rather a key strength.

Working within clearly defined permissions
The agent only accesses the information for which it has been explicitly authorized. It only sees what is relevant to the specific use case in the warehouse—such as inventory, storage locations, or availability. Which data is available can be controlled in detail and adjusted at any time.

Not all information is – or should be – available
A warehouse chatbot does not need to know everything. Sensitive data such as purchase prices or calculations should be deliberately excluded. This ensures that access remains secure, controlled, and role-based. The agent thus becomes a reliable helper in the warehouse – not an uncontrollable data information system.

Deliberately not a universal AI all-rounder
The warehouse agent is not designed to answer general knowledge questions or generate texts. For example, it does not know any sights in Paris, but it knows exactly how many chairs labeled “Paris” are available in the warehouse, what variants exist, where they are located, and whether the quantity is sufficient for the next order. This technical focus ensures precise answers, short dialogs, and a high degree of reliability in day-to-day business.

Not a replacement for specialist processes, but targeted support
The chatbot does not replace warehouse processes or the expertise of employees. It provides support where quick information is needed – and fits into existing workflows and processes. Decisions, bookings, and process responsibility remain in the system and with the people.