New to process mining

If you have never done process mining, this track takes about an hour and assumes nothing. By the end you will have discovered a process model from real event data, read it, found the slowest step, and know what to do with a log of your own.

The idea in one paragraph

Most work leaves a trail. Every time someone approves an order, ships a parcel or closes a ticket, some system writes a row with a timestamp. Process mining takes those rows and reconstructs what the process actually looks like — not the flowchart somebody drew in a workshop, but the paths real cases took, including the loops, the rework and the shortcuts. You do not model the process; you discover it.

The track

1. What an event log is — three columns, and why they are enough.

2. Your first process map — import the demo log and discover a model.

3. Reading a discovered model — what the boxes, arrows and numbers mean, and which parts to distrust.

4. Where the time goes — turning a map of what happens into a map of what it costs.

5. Bringing your own data — the shape your export needs, and the three columns to look for.

You do not need to install anything

Promenade runs in the browser tab you already have open, and the demo logs ship with it. Nothing is uploaded: the log is parsed and queried on your own machine.

After this track

If you want to go deeper into the algorithms — why one miner produces a different model from another, what “sound” means, how conformance is computed — the expert track picks up exactly there. If you would rather turn this into an analysis someone at work will act on, go to the practitioner track.