Agentic AI Reference Architecture

Cognition · Software component

Counterfactual Explainer

Software componentCognitionCognition & Memoryarc:CounterfactualExplainer

An explanation component that identifies which fact values would need to change, relative to rule condition thresholds, for a rule-based decision to come out differently.

Responsibility. Explains decision boundaries by stating the minimal changes that would alter the outcome.

Also known as: Decision boundary explainer, Counterfactual explanation, What-would-need-to-change explanation, Contrastive explanation generator, Diverse counterfactual explanations

Variant of Decision Explainer abstract

When to choose. Choose when users need to know the minimal changes that would flip the decision (why-not / recourse).

sends data toinvokesis invoked byalternative toreadsspecializessends data toalternative toalternative tois routed to byreadsExplanation Presenter: sends data toExplanation PresenterDecision Engine: invokesDecision EngineRule-Based Decision Engine: is invoked byRule-Based Decision EngineDecision Factor Explainer: alternative toDecision Factor ExplainerProduction Rule Base: readsProduction Rule BaseDecision Explainer: specializesDecision ExplainerSummary Explanation View: sends data toSummary Explanation ViewExample-Based Explainer: alternative toExample-Based ExplainerLocal Surrogate Explainer: alternative toLocal Surrogate ExplainerExplanation Method Selector: is routed to byExplanation Method Selec…Rule Firing Trace: readsRule Firing Trace
Direct neighbourhood (hover for relationship types)

Relationships

invokes dependency

is invoked by dependency

reads dependency

is routed to by dynamic

sends data to dynamic

alternative to variability

Design guidance

Quantitative guidance

As stated by the sources; verify before use.

Classification

Patterns
DiCE (Diverse Counterfactual Explanations)Contrastive explanation
Technologies
DiCE
Quality attributes
Explainability (NIST AI RMF: explainable and interpretable)Functional suitability: correctness and validity (ISO/IEC 25010 | NIST AI RMF: valid)
Frameworks & regulations
Fair Credit Reporting Act (specific reasons for credit denial)

Sources

  1. Ch5.11: T. Nguyen, "Rule-Based Decision Making Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.11. ISBN: 9798244538229.
  2. Ch9.3: T. Nguyen, "Sandboxing and Transparency Foundations," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 9.3. ISBN: 9798244538229.
  3. Ch10.5: T. Nguyen, "Human-over-the-Loop," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 10.5. ISBN: 9798244538229.
  4. Ref10.02: "Explainability and Interpretability in Agent Systems," unpublished reference note (02-Explainability-Interpretability.md), Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam supplementary materials, 2026. unpublished note
  5. Ref10.07: "Chapter 10 Summary: Human-AI Interaction and Oversight," unpublished reference note (07-Chapter-10-Summary.md), Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam supplementary materials, 2026. unpublished note