NFR Repeat-Finding Detector
Description
Identify recurring themes across Notices of Findings and Recommendations and link them to root causes and responsible systems/process owners.
AI / analytics pattern
NLP clustering + knowledge graph
Automation level / stage
assistive/analytics with human review
Expected benefit
Improves CAP targeting and reduces repeat findings.
Audit / financial statement impact
Supports material weakness remediation and audit opinion progress.
Controls / human review
Human review for exceptions and recommendations; maintain evidence packages, lineage, source citations, model cards, data-quality checks, and periodic QA sampling.
Data needed
source transactions, reference tables, audit evidence, user/action logs as applicable
Possible metrics
cycle time, match rate, exception aging, NFR closure, audit support effort, data quality score
MVP scope
Pilot with one reporting entity, one fiscal quarter, and read-only outputs before workflow integration.
Related material weakness / control objective
Directly tied to cited DoD FM/audit priority or system mission.