Workflow Inspector Signal · Changes worth investigating
Spot a measured change. Keep the cause open.
Signal compares a baseline with a later, non-overlapping window. It surfaces material increases in recorded exceptions, rework, ownership gaps, or handoff latency so your team can choose what to investigate.
Open Signal in your workspace · Compare SaaS workspace plans
Sign in, then choose an eligible SaaS Onboarding workspace and workflow. Product access is checked by the server.
What changed in our recorded workflow?
Measured comparison
See the earlier and later metric values, sample support, and the rule that identified a change. A source gap or small sample stays part of the result.
Evidence beside interpretation
Inspect the supporting event references separately from the possible explanation and suggested investigation. A change does not establish its cause.
Saved review history
Save a run and append a human review of a signal. Review status records a person's assessment; it does not prove that a condition has been fixed.
Illustrative example
Fictional example, not customer evidence or a measured customer outcome.
- Observation
- A fictional workflow's completed, matched handoffs have a median of 4 hours in the baseline window and 7 hours in the later window.
- Interpretation
- The observed handoff duration increased. This alone does not show whether capacity, routing, missing context, or a different case mix caused the change.
- Next step
- Inspect the referenced handoffs, record a review, and collect evidence before changing the process.
What you need
- Accepted operational events for one workflow, with usable timestamps and the event types needed by the metric.
- A baseline window and a later, non-overlapping comparison window, both ending no later than the present.
- Enough comparable samples to evaluate a rule. Unsupported comparisons remain unavailable instead of producing a signal.
Known limits
- A rule-based change detector, not a prediction of future failure or a causal diagnosis.
- No background alerts or automatic interventions are included in this version; run and review comparisons in the workspace.
- Review labels and evidence-strength categories are not calibrated probabilities, and no employee ranking is produced.
Results describe the accepted records in the selected time window. Missing timestamps, ambiguous event pairs, unavailable metrics, and unverified source completeness remain visible limitations. Missing data is not treated as zero or filled in with inferred activity.
Workspace access and privacy
Sign-in and membership in the selected workspace are required. The server checks workspace membership, product capability, and permissions for each request. Authorized retained history can remain readable after a subscription ends; new records require an active eligible subscription. Evidence and saved records stay scoped to that workspace and workflow; public product pages contain no customer records.
Privacy policy · Security information
Review current plan availability · Workspace subscription terms
Subscriptions renew monthly until canceled under the workspace terms. Your team remains responsible for decisions and implementation.
Questions about workspace fit? Email Walter.