Demo · made-up data
Know when something breaks. Before your customer does.
Automations rarely fail loudly. They stop, and nobody notices for a week. This is how we watch them: step through one day and see what you would be told, and when.
Step 1 of 7 · 08:00
Just connected
Nothing has reported yet. We don’t call that working. It stays “not confirmed” until the first real signal arrives.
Status at 08:00
3 of 3 not confirmed yet.
- Not confirmedWebsite enquiries → CRMNo signal yet
- Not confirmedInvoices → accountingNo signal yet
- Not confirmedDaily backupNo signal yet
What you would be told
Nothing. No news is sent when there is no news.
Incident log
No incidents so far.
Four rules behind it
- Unknown is not healthy
- A workflow shows as working only after it proves it. Until then it says “not confirmed”.
- Silence is a failure
- Every workflow has a deadline for its next signal. Missing it raises a problem, even with no error anywhere.
- One alert per problem
- A workflow that fails twenty times in a row is one incident and one message.
- Recovery is confirmed
- You hear when it works again, and how long it was down. Every incident stays on the record.