Large Language Models can accomplish Business Process Management Tasks

Within the realm of Business Process Management (BPM), the incorporation of LLMs is redefining the management and analysis of processes. This paper demonstrates that even without tailored solutions, LLMs proficiently execute an array of BPM tasks, such as extracting information from textual descriptions and evaluating process tasks for automation potential. The findings suggest that LLMs can serve as versatile, all-in-one tools in the BPM toolbox.

• *Demonstrates LLMs’ versatility in text-related BPM tasks*
• *LLMs successfully mine both imperative and declarative process models*
• *Equally adept at assessing task suitability for robotic process automation*
• *Performance comparable to specialized BPM solutions without extensive configuration*
• *Insight into LLMs as a general-purpose instrument in the BPM landscape*

The implications of LLMs in BPM, as illustrated by these experiments, highlight the significance of AI in streamlining organizational processes and fostering research that connects BPM and AI technologies.