Scheduled is the one place that shows everything running on a timer — reports that run on a rhythm (every morning, every Monday, month-end) and recurring analyses such as a monthly cash-runway study. For each one you see how often it runs, when it last ran, whether that run worked, and when it is due next.
- Each run uses the report exactly as published at that moment — a report you've since improved runs in its improved form.
- Times and due dates are South African time.
What each result means
The last run has one of four results:
- OK — the run finished and the sources it depends on moved as expected.
- Stale — the run finished, but at least one source has not moved since the previous run. Treat the number you are looking at as a repeat of older data until the named source pulls again; it is not a fresh measure of today.
- Short — the run finished but stopped at a row limit, so the figures cover only the period or rows named on the screen.
- Failed — the run did not finish. The message stays brief and does not show private technical details; if support needs more, they use separate diagnostics.
Failed, stale and short results are not just page badges: the report owner is told in the inbox and by email.
Setting something up
Schedules are created where the work is: ask the AI analyst to run a report on a schedule, or set a recurring analysis going from the study itself. Once it exists, it appears here.
Changing what runs
Administrators can change any of the company's schedules from this screen — not only the ones they set up themselves, so a schedule doesn't become unmanageable when the person who created it moves on. Everyone else can see the list.
- How often — switch between every day, every week, every month, or month end.
- Pause — stops the clock and keeps everything, including past results. Resume any time. A paused item stays on the list; it doesn't disappear.
- Run now — runs it once, immediately. This does not use up the scheduled run: a monthly pack run by hand on the 3rd still runs automatically on its normal date.
- Remove (reports) — stops the report running, permanently, and its saved run history goes with it. If you only want it to stop for now, pause it instead.
- Stop (analyses) — takes the study off its timer and leaves the study itself intact, with its frozen base, confirmed assumptions and past results.
Running a recurring analysis by hand still goes through the usual assumption confirmation — if assumptions are waiting for you, it says so instead of computing on unconfirmed inputs.