Why Kea exists.

Kea is a desktop application for reading an event log properly. It groups cases by how they actually ran, draws the process as a graph from the log alone, puts the individual cases on a timeline, charts their attributes, and tells you in numbers what separates one group from another.

Every panel shares a selection, so a click in one narrows all the others to the same cases. Every step is recorded in a tree you can walk back through or hand to a colleague. That is the whole idea: from a pattern to the cases behind it, without losing the trail.

The logs worth studying are the ones you cannot upload: patient pathways, claims, customer orders, internal tickets. Every cloud process-mining product starts by asking you to send that data somewhere. Kea does not, because a desktop application never has to.

It costs us some conveniences. There is no shared workspace in a browser and no dashboard for the board. In exchange, the question of where the data went has a one-word answer.

Kea does not run your process or change it, so it is not a workflow engine or a BPM suite. It has no connectors into your systems: you export a log and Kea reads it. And the language-model features are optional and local, so nothing in the analysis depends on them.

It is a tool for the person who has to explain why the numbers look the way they do, and for the person they have to explain it to.