The idea of a company-specific ChatGPT that answers the operator's questions about production quickly and proactively is compelling:
Such an application would significantly shorten the time required to solve operational challenges in the tobacco industry’s production processes. As a result, efficiency and quality of production processes increase while costs decrease. "What sounds like a utopia, we are already providing to companies today. Together with our partner company DAIN Studios, we have developed a GenAI tool to efficiently utilize customer-owned knowledge management systems," reports Alexander Pfaffenrot, Senior Product Owner at Körber. "As a leading provider in the tobacco industry, we incorporate our in-depth knowledge of machines, their operation, and maintenance into customer-specific solutions. DAIN Studios provides the technological foundation for the tool."
News & Stories - 09/11/2025
ChatGPT for the Shopfloor
Customized ChatGPT based on the entire company knowledge
Our customer-specific knowledge management system forms the data foundation that the KMS Co-Pilot accesses. This enables significant advantages in day-to-day production, especially in increasing MTBF (Mean Time Between Failures). Whereas in the past, finding solutions to questions and issues in production required a time-consuming, often inefficient, and sometimes unsuccessful manual search, the operator now receives optimized, tailored support directly in the production environment, based on the customer-specific knowledge management system.
"This works only because our solution is not off-the-shelf but 100% customer-specific. There is nothing comparable on the market so far," explains Alexander Pfaffenrot. "The chat responses of the KMS Co-Pilot are of high quality: step-by-step instructions, links to sources for additional information, and extended solution offerings that integrate areas such as purchasing, spare parts, or training. It is a learning system that is continuously refined. The results are evaluated by experts in terms of accuracy, completeness, and usefulness."
Secure and Future-Proof
The KMS Co-Pilot is a future-proof tool for our customers that continuously supports their development. "New information in the KMS and new insights at Körber can be continuously incorporated, keeping the answers in the KMS Co-Pilot up to date," says Pfaffenrot. "Since the information used is valuable company property, we focused on data security from the beginning: Our KMS Co-Pilot exclusively accesses the customer-specific data base and not the internet. By this separation from the World Wide Web, all data remains secure within the customer's system."
Three Questions for Alexander Pfaffenrot, who has significantly co-developed the KMS Co-Pilot as Senior Product Owner.
The KMS Co-Pilot is a sort of all-purpose tool for challenges at the machine. Where does the tool source its "knowledge" from?
Alexander Pfaffenrot: In the implementation phase, we train the KMS Co-Pilot by connecting it to the customer's KMS and supplementing the information with our expertise. This includes technical and product specifications, as well as guides, documentation, and user manuals from the customer's KMS, which are then enhanced by our Körber solution expertise.
Can the tool be imagined as a kind of ChatGPT for all questions related to machine handling?
Alexander Pfaffenrot: Of course, it is a continuously evolving and self-learning tool, but essentially, the comparison is correct. The KMS Co-Pilot is a chatbot that, for example, provides instructions for part replacement questions immediately. Through the further development of the tool, it is conceivable that the KMS Co-Pilot, if necessary, could trigger purchase orders and check the availability in the warehouse. The great advantage: Our system connects the entire company knowledge on this topic and creates a seamless framework. Directly on the shop floor, it provides fast and easy access to all the information that used to be time-consuming to gather and possibly did not offer the optimal solution. Processes that, in daily operations, the operator often simply doesn't have the capacity for, and which cost valuable time in production. Through the IWS approach (IWS = Integrated Work System), which shifts responsibility to the shop floor, our tool enables key users like operators, shift planners, and production managers to make informed decisions quickly and directly on site.
What customer challenges does the KMS Co-Pilot aim to address as a solution?
Alexander Pfaffenrot: The key performance indicator here is the Mean Time Between Failures. The main goal of using the KMS Co-Pilot is to increase this key value. We can efficiently tackle this central challenge of our customers by making their existing data from the KMS easily accessible, networking it, combining it with our know-how, and optimizing it for decision-making processes on the shop floor.
Share this article