Resolving Alerts

The Impact Cascade Engine doesn’t just tell you what broke — it tells you how to fix it with specific suggested actions.

Suggested Actions

Every cascade alert includes one or more suggested fixes. These are specific, actionable recommendations:

Examples of Suggested Actions

Cascade alerts are created when a controlled document is published, and they’re routed to the owners of linked documents:

AlertSuggested Action
”Linked SOP ‘Customer Refunds’ republished (v3)""Review the Work Instruction you own for steps affected by the SOP change"
"Parent Policy ‘Data Privacy’ republished""Confirm your SOP still complies with the updated policy requirements"
"Work Instruction your SOP depends on was republished""Verify the referenced steps and update your SOP if needed"
"Training doc republished — acknowledgements now stale""Re-assign Required Reading so assignees acknowledge the new version”

Working Through an Alert

Acknowledging an Alert

When you’ve seen an alert and are working on it:

  1. Open the cascade alert
  2. Review the suggested action and the AI explanation
  3. Click “Acknowledge” — the alert stays open, but everyone can see it’s being handled

Manual Resolution

For complex situations requiring judgment:

  1. Open the cascade alert
  2. Review the impact chain
  3. Take your own corrective action (update the affected document, re-assign Required Reading, etc.)
  4. Click “Resolve” to close the alert

Dismissing Alerts

For alerts that don’t require action:

  1. Open the alert
  2. Click “Dismiss” with an optional reason
  3. The alert is removed from your active list

The Impact Summary View

For large cascades affecting many entities, use the Impact Summary View:

  • Severity breakdown — how many critical, warning, and info alerts
  • Entity breakdown — which documents and owners are affected
  • Batch actions — acknowledge, resolve, or dismiss multiple alerts at once
  • Audit trail — see who resolved what and when

Best Practices

  1. Prioritize by severity — critical first, then warnings, then info
  2. Review before applying — suggested actions are smart but verify they make sense for your context
  3. Use batch resolution when multiple alerts have the same fix
  4. Check the audit trail to ensure nothing was missed

The goal: When you make a change in BFF, you should be able to see and resolve all downstream effects within minutes — not discover broken processes days later.