GSA Logo - U.S. General Services Administration

USA Hire℠ Agentic Platform

Closed-Loop Validity

Step 1 • Job Analysis Agent

A defensible job analysis for any role — even ones that don't exist yet.

Describe a new role in plain English. The agent uses Critical Incident Technique to extract tasks, maps them to the federal KSAO taxonomy + O*NET, and classifies each KSAO as covered, adaptable, or novel — Uniform Guidelines §14C(2) compliant.

Working example

AI Governance Officer — net-new role, no precedent series

Manager intake started 2026-05-12 09:42 ET · Draft v3

Manager intake (CIT-guided)

Step 2 of 4
We need to hire an AI Governance Officer. The role oversees responsible AI deployment across the agency, audits LLM systems for bias, and partners with privacy and security teams. No existing series fits cleanly.
Got it. Walking through Critical Incident Technique. Tell me about a recent moment where this role would have made a real difference if it had existed.
Last quarter our chatbot pilot leaked draft policy text in completions. Nobody owned the review process. We needed someone who could redesign the eval suite, brief leadership in plain language, and codify the policy.
That gives me three task families: governance design, technical evaluation of AI systems, and executive communication. Extracting 24 task statements now and mapping to the federal KSAO taxonomy…

Task statements

Extracted by the agent · pending SME review

  • Design AI governance frameworks aligned with NIST AI RMF and OMB M-24-10
  • Conduct bias and fairness audits of agency LLM deployments using standardized eval suites
  • Translate technical AI risk findings into plain-language briefings for executive leadership
  • Partner with CISO and Privacy Officer on cross-cutting AI risk reviews
  • Define and maintain agency-specific AI acceptable-use policy
  • Review vendor AI products against federal procurement and ATO requirements
  • Advise program offices on responsible AI deployment decisions

+ 17 more task statements extracted

KSAO overlap report

Similarity check against federal PD corpus + O*NET

Covered

7

Adaptable

1

Novel

3

KSAO portfolio

Weights derived from task criticality × frequency ratings

Federal AI policy (NIST AI RMF, OMB M-24-10, EO 14110)

novel

0.95

AI/ML system architecture fundamentals

adaptable

0.85

Cybersecurity & FISMA frameworks

covered

0.78

Privacy law (Privacy Act, SORN process)

covered

0.72

LLM bias and fairness evaluation

novel

0.92

Technical writing for executive audiences

covered

0.88

Cross-functional facilitation

covered

0.80

Vendor risk assessment

covered

0.70

Complex systems reasoning

covered

0.85

Integrity & ethical judgment

covered

0.92

Comfort with technical-policy ambiguity

novel

0.78

Similar historical roles

2210 IT Specialist (INFOSEC) — GS-13/14

62% overlap

Strong on cyber, FISMA, vendor risk; weak on AI/ML evaluation.

0301 Program Analyst (Privacy) — GS-13

48% overlap

Strong on privacy/SORN; partial on cross-functional facilitation.

1550 Computer Scientist (Research) — GS-14

41% overlap

Strong on AI/ML fundamentals; weak on governance & policy.

0132 Intelligence Specialist — GS-13

28% overlap

Adjacent on analytic rigor; limited overlap on technical AI specifics.

Novel KSAOs requiring new assessment development

These KSAOs have no validated counterpart in the existing catalog. They route automatically to the Assessment Design Agent (Step 2) for fresh content development.

Federal AI policy (NIST AI RMF, OMB M-24-10, EO 14110)

No prior validated assessment exists. Requires fresh item development and SME review (Hunt + Quintela).

LLM bias and fairness evaluation

Build live evaluation simulation: applicant runs a fairness audit on a synthetic LLM output set.

Comfort with technical-policy ambiguity

Adapt SJT items from senior policy roles; new content needed for AI-specific dilemmas.

Hand off to the Assessment Design Agent

The job analysis is ready. 7 KSAOs reuse existing assessments, 1 needs transportability, 3 need fresh development. The Design Agent will optimize a battery covering all 11 with predicted validity, time, and adverse impact constraints.

Continue to Step 2