Large language models (LLMs) have significant potential to support companies—especially SMEs—in capturing and improving their process models (BPMN). They can take on the role of a digital process modeler, create process models based on dialogue, and analyze and further develop existing models as chatbot-supported modeling copilots.
In the KICoPro project, DFKI is developing an innovative modeling tool based on finely tuned open-source LLMs that enables process modeling and optimization in natural language. The modeling copilot to be developed can automatically convert textual descriptions into business process models according to the BPMN standard and systematically supports companies in the structured collection of their process flows.
Based on its trained process knowledge, the KICoPro copilot can also generate well-founded optimization suggestions, for example by recommending improved process steps in comparison with established reference processes. This not only simplifies modeling, but also promotes continuous process improvement.
The solution is particularly beneficial for industries that are undergoing an ongoing digitalization process, such as municipal utilities and small and medium-sized manufacturing companies. KICoPro thus makes a concrete contribution to digital transformation by significantly facilitating access to professional business process management.
Contact:
Alexander Rombach | Alexander_Michael.Rombach@dfki.de
To Product Page