Executive artifact

AI/Human Review Audit Trail

Show where Companion assisted and where humans reviewed or overrode.

Audit sampleRisk / Governance / CHROCompanion Intelligence

Sample data boundary: deterministic representative data only. No client records, live market survey data, or protected scoring and derivation mechanics are exposed.

Back to artifacts Request proof access

Printable artifact shell

Board/executive structure.

Risk / Governance / CHRO

AI/Human Review Audit Trail

Decision supported: Show where Companion assisted and where humans reviewed or overrode.

Companion prompt

Representative section showing evidence, reviewer posture, caveats, and next action. Anonymized sample content only.

Advisory output

Representative section showing evidence, reviewer posture, caveats, and next action. Anonymized sample content only.

Reviewer action

Representative section showing evidence, reviewer posture, caveats, and next action. Anonymized sample content only.

Override log

Representative section showing evidence, reviewer posture, caveats, and next action. Anonymized sample content only.

Audit metadata

Governance, approval, and Companion advisory note.

Governance status: Audit sample
Human boundary: AI observations remain advisory and logged.
Audit metadata: Mode, context hash, reviewer, override status, timestamp.
Companion note: Companion flagged a gap; reviewer accepted the gap as a condition.

Embedded Companion

The AI surface challenges and explains; humans approve.

Companion IntelligenceAI prepares · consultant reviews · client decides

Governed prompts inside the workflow.

Evidence gap: market source freshness requires reviewer confirmation.

Assumption to validate: attrition rate sensitivity exceeds budget tolerance.

Human review required before final recommendation.

Alternative scenario: 50th percentile positioning with targeted critical-role exceptions.

Cannot approve · Cannot finalize · Cannot fabricate market data · Cannot bypass human review