Agentic AI Reference Architecture

Observability & Evaluation · Software component

Reasoning Chain Decomposer

Software componentObservability & EvaluationObservability & Evaluationarc:ReasoningChainDecomposer

An evaluation preprocessing component that splits a reasoning trace into steps and each step into Reasoning Content Units labelled as premises or conclusions.

Responsibility. Decomposes reasoning traces into independently verifiable content units.

Also known as: RCU decomposer, Reasoning content unit extractor

is orchestrated bysends data toreceives data fromsends data tosends data tosends data tois configured byis configured byEvaluation Harness: is orchestrated byEvaluation HarnessReasoning Consistency Checker: sends data toReasoning Consistency Ch…Evaluation Trace Sampler: receives data fromEvaluation Trace SamplerEntailment Step Validator: sends data toEntailment Step ValidatorGoal Alignment Checker: sends data toGoal Alignment CheckerInformation Gain Scorer: sends data toInformation Gain ScorerTagged Reasoning Template: is configured byTagged Reasoning TemplateReasoning Trace Schema: is configured byReasoning Trace Schema
Direct neighbourhood (hover for relationship types)

Relationships

is configured by structural

receives data from dynamic

sends data to dynamic

is orchestrated by control

Classification

Patterns
Reasoning Content Unit (RCU) decompositionRECEVAL correctness and informativeness framework
Risks mitigated
Conflating distinct reasoning failure modes in monolithic assessment

Sources

  1. Ch3.9: T. Nguyen, "Reasoning Quality," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 3.9. ISBN: 9798244538229.