Traces
Did it run the way it should, every time?
A Trace follows each real run of a process through the records it leaves behind: every order, every close, every leaver. It judges each run against the path you expect, step by step, with nothing created or changed.
Trace · last 30 days, 50 leavers
Every leaver loses access the same day
Termination recorded50
Okta account deactivated1 late49
Microsoft 365 license removed1 missed48Laptop wiped or returned2 not observed46Each step expected within 8 hours of the termination
Recent leavers
J. Alvarez: Okta off at 09:40, Microsoft 365 license still assigned 6 h later.
Verdicts
Every run gets a verdict.
Not just the ones that failed.
Each run of the process is judged against its expected path and lands in exactly one of these.
Completed
Every expected step happened, in order, inside its time window.
Late
Every step happened, but at least one landed outside its window.
Deviated or missed
A step happened out of order, or never happened. The step and the record are named.
Not observed
Panaptico could not read the records for a step. Shown as its own verdict, never as complete.
How a Trace works
Read the records the process leaves.
One run per real event.
Traces are read-only. They never create test data; they judge what actually happened.
- 1Start
Pick the starting event
A closed deal, a termination, a month-end trigger. Each one starts a run.
- 2Expect
Describe the path
The steps that should follow, in which systems, and how long each may take. Branches like one of are supported.
- 3Follow
Match each run
Panaptico reads each system's own records and lines them up against the path, run by run.
- 4Judge
Verdict per run
Counts per step show exactly where runs fall off, and each deviated run links to its records.
Why it matters
Averages hide the one that broke.
Traces don't.
Every run, not a sample
Fifty runs means fifty verdicts. The one that deviated is named, not smoothed into a rate.
Where runs fall off
Counts at each step show the exact handoff where the process leaks.
Evidence per run
Each verdict links to the records in each system that support it.
Examples
Traces teams run.
For processes that run many times.
Works with
The rest of the platform.
One record of what's true.
Know every run held.
Not just the ones you looked at.
Pick a process that runs many times a week. Panaptico follows each run through its own records.