Skip to Content
OperationsIncident response

Incident response

The Nebbos platform operates a detection-through-postmortem incident-response procedure supported by the substrate audit chain. Every incident is recorded as a distinct chain entry from declaration through publication of the postmortem.

Severity classification

SeverityImpactNotification
Sev 1Substrate unreachable, audit-chain integrity failure, or platform-wide data-access failureAffected operators paged; status board declares within five minutes of declaration
Sev 2Functional-surface impairment; core read and write paths remain operationalAffected operators notified within one hour of declaration; status board within thirty minutes
Sev 3Localised performance degradation; documented workarounds availableStatus board note within two hours of declaration
Sev 4Enquiry or feature request; not an incident, tracked in the same queue for measurabilityStandard support response cadence

Detection

Synthetic-probe metrics are emitted every sixty seconds from three geographic regions. Continuous audit-chain integrity checks execute in parallel. The on-call rotation is paged within two minutes of a Sev-1 or Sev-2 trigger, in advance of operator-observable impact.

Operator-reported incidents are opened via enterprise@nebbos.ai with the subject line prefixed [incident]. Response-time commitments for these reports match the SLA.

Communication

Every Sev-1 and Sev-2 incident receives:

  • A public status-board entry within five minutes of declaration
  • Direct email to every affected enterprise operator within one hour, naming affected surfaces and next-update expectations
  • Recurring updates every thirty minutes for Sev-1 and hourly for Sev-2 until mitigation is confirmed; each update names the change since the previous update
  • A resolution notice within twenty-four hours of mitigation, naming the corrective actions committed

Postmortem

Every Sev-1 and Sev-2 incident is followed by a postmortem published within seven days of resolution.

  • Contents: incident narrative, affected surfaces and operators, root-cause analysis (five-whys), corrective actions already applied, corrective actions committed with named owners and dates.
  • Approach: blameless.
  • Publication surface: status.nebbos.ai .

Postmortems are governance artefacts. Corrective actions are recorded in the governance corpus and, where applicable, as architectural decision records. Postmortems constitute part of the post-market monitoring plan referenced in the EU AI Act Annex IV pack.

Audit-chain entries

Every incident produces the following audit-chain entries: declaration, each communication update, mitigation confirmation, resolution, and postmortem publication. The audit chain provides an inspector-walkable record of every incident lifecycle without requiring further coordination with Nebbos.