This master's thesis is carried out in cooperation with Stellantis. The aim is to link the company's existing structured process models with unstructured, textual process descriptions.

First, a suitable exchange format must be found to make the extensive process models known to a language model.

The next step is to investigate how validation and optimization proposals for process models can be created by incorporating the process descriptions.

The prototypical implementation of an LLM agent that can fulfill the aforementioned tasks is to be developed.