Model Validation Agent
Runs independent model validation analyses and drafts the validation report.
Reproduces the model developer's results, runs the benchmarking and outcomes analysis, probes the assumptions, and drafts the validation report to SR 11-7 structure. Spawns parallel test sub-agents for replication, benchmarking and sensitivity; an independent agent re-derives the results as a check, and the accountable model-risk officer owns the sign-off.
Authority
Recommend for approval
Team role
Coordinates the work
Handoffs
Named collaborators
The role
What it owns and where its authority ends
Desk
Model Risk Management
Desk workflow
Model inventory, then validation (conceptual soundness, outcomes analysis, benchmarking), then ongoing performance monitoring, then documentation and annual revalidation.
Collaboration
Coordinates specialist contributions
Decision boundary
An accountable reviewer commits the decision or action.
Systems and capabilities involved
Model inventory + documentation
Validation test harness
replication, benchmarking
Model + data environment
Independent validation re-derivation
results re-derived by a second agent; sign-off stays with the accountable model-risk officer
Handoffs
What this role gives and receives
Capabilities offered
The handoffs name the next owner or specialist and the work that moves between them.
Handoff to
External handoff
Every model-owning desk in the bank (incl. the AI agents here)
External handoff
Financial Crime (Scenario Tuning) for tuning-model sign-off
Context
What the role needs to do the work
- Current work
- The model under validation, its documentation, and test results so far.
- Prior interactions
- Prior validations of similar models and recurring findings.
- Policies and reference
- SR 11-7 expectations, validation methodology, the model inventory.
- Working method
- Validation-test playbooks per model class.
Illustrative workflow
How the work moves
Starting point
Annual revalidation due on the retail PD scorecard.
- 01
Spawn sub-agents: replicate the developer's results, benchmark, run sensitivity tests.
- 02
Probe assumptions and check outcomes analysis against realized defaults.
- 03
Identify findings and rate their severity.
- 04
Draft the validation report to SR 11-7 structure with cited evidence.
Result
A draft validation report with two medium findings (a data-quality gap and a stale segment), routed for an independent re-derivation check and the accountable model-risk officer's sign-off.
Checks and boundaries
What must be tested or reviewed
- 01Every validation requires an independent re-derivation check and sign-off by the accountable model-risk officer (SR 11-7).
- 02Test reproducibility checked; results traced to the model environment.
- 03Agent-as-judge review of report completeness vs. the validation standard.
- 04Independence guardrail: cannot validate a model it helped develop.
Human authority
An accountable reviewer commits the decision or action.
Keep exploring