When the documentation says what actually happens
Documented target processes and actual process behaviour had drifted apart — without one shared view, every analysis stayed contestable.
Starting point
The workflows were documented. They just didn’t reliably describe what actually happened in operations — between process intent and operational reality sat a gap no one could put a number on. As long as the two stood side by side, any analysis could be dismissed with the objection that it measured the wrong picture.
What made it harder was that the two halves of the truth sat in different tools: what the IT systems revealed about the real sequence was locked in process mining. What the workforce knew about their own work appeared nowhere in it.
Approach
Rather than choosing one side over the other, the two were brought together. The process-mining results were carried over and enriched with the knowledge in people’s heads — data from the systems on one side, experience from the workforce on the other.
To do that, the lived actual processes were identified first, with AI support instead of being gathered again in workshops. They were then harmonized with the already documented workflows. That is what made the differences between what was intended and what happened nameable in the first place.
Outcome
Target and actual now show the same view — a 360° picture of the processes that neither source could have produced on its own. That isn’t an end in itself: only on this basis does a process analysis carry weight. It then assesses the workflow that actually runs, rather than the one that was written down once.
“I’m a process manager trying to mow the field with a scythe. And the AI is the combine harvester I’d been missing.”
Lisa Brückner, Head of Process Management · Electrical engineering