Processes — write one, run it

A process is the thing you explain to every new colleague: how a customer is onboarded, how a month is closed, how a job is handed over. Written once, it stops being folklore.

Two tabs: Processes — the definitions — and Runs — every time one was carried out.

Writing one

A process has sections and numbered steps. A step is one instruction a person could follow: "Create the customer folder", "Copy the templates from /vorlagen", "Tell the team".

Write it the way you would say it out loud. Steps that are too clever to follow are steps nobody follows.

Running one

Start run makes a run: a checklist of that process, with its own progress. Tick steps off as you go, reopen one you ticked too early, and press Complete run at the end.

Runs stay in the list when they are finished — the list says In progress or Completed, grouped by the process they came from, with who started it.

Runs that start themselves

A process can be bound to a record, so creating that kind of record sets it off. A new customer starts the onboarding run, and the run names the customer it is about.

That is where the initiator column earns itself: it says whether a person or the agent started it, and a failed automatic run says why, right there in the list — because nobody was watching at the moment it went wrong.

Letting the agent run it

"Start the onboarding process for Meyer GmbH" creates the run. The agent can also work the steps and tick them off as it does them — it sees the same numbered list you do.

Steps with consequences are approved like anything else, see when the agent asks first.