AI Event Intelligence and Governed Remediation

Durable anomaly detection, cross-service correlation, root-cause context, and operator-approved remediation in QNSI.

QNSI turns retained AI and platform events into operational evidence. The workflow is designed for accountable security operations: it can identify and correlate signals, prepare root-cause context, and propose remediation, but it does not bypass tenant authorization or configured operator approval.

What is available

  • Durable event retention for AI and platform signals used by the intelligence workflow
  • Anomaly detection across retained events
  • Cross-service correlation that groups related signals into incidents
  • Root-cause context generated from anomaly, correlation, and error-pattern evidence
  • Remediation proposals that remain pending until an authorized operator approves or rejects them
  • Cryptographic verification for signed dashboard evidence where enabled

The Cloud Portal exposes dedicated views for the intelligence dashboard, anomalies, correlation, root-cause analysis, prompt security, bias, costs, compliance, scaling, errors, self-healing, and remediation. Availability is controlled by the tenant's plan, feature flags, identity, and policy.

Approval boundary

An AI recommendation is not a production change. QNSI keeps proposal generation separate from approval and execution so the audit trail can show:

  1. which retained evidence caused the proposal;
  2. which root-cause context was presented;
  3. who approved or rejected the proposal; and
  4. which resulting action was attempted and observed.

This separation supports reviewable automation without turning model output into an unaudited administrative action.