Agentic AI Reference Architecture

Human Oversight · Data store

Override Rationale Log

Data storeHuman OversightExperience & Human Oversightarc:OverrideRationaleLog

A store of human-stated reasons accompanying overrides, approvals despite flags, and initiative takeovers, capturing factors the agent's training data did not emphasize.

Responsibility. Records why humans overrode or redirected agent decisions.

Also known as: Human rationale record, Bidirectional transparency log

is read byis written byis read byis read byData Curator: is read byData CuratorApproval Review Console: is written byApproval Review ConsoleHandoff Context Packager: is read byHandoff Context PackagerOverride Pattern Analyzer: is read byOverride Pattern Analyzer
Direct neighbourhood (hover for relationship types)

Relationships

is read by dependency

is written by dependency

Design guidance

Classification

Patterns
Bidirectional transparency
Risks mitigated
Incompatible human-AI mental modelsHuman overrides treated as arbitrary

Sources

  1. Ch10.2: T. Nguyen, "Proactive Agents," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 10.2. ISBN: 9798244538229.