The carrier moved every model — statistical and AI alike — into one SR 11-7 register with independent validation, backtesting and continuous drift monitoring, extended with bias and explainability controls for AI/ML, all on a tamper-evident audit trail.

A large multiline P&C carrier tracked its pricing, reserving and growing roster of AI/ML models across spreadsheets and email. Validation status was hard to evidence, drift went unnoticed between annual reviews, and the model risk team could not answer examiners quickly on which models fed regulatory numbers.

Book a demo All customer stories

Illustrative outcome

100%

of models carrying a current validation status

Model Risk & AI Governance
Risk Management
Compliance

Illustrative and capability-framed — representative of platform capability, not a verified named-customer claim.

The challenge

The solution

The carrier moved every model — statistical and AI alike — into one SR 11-7 register with independent validation, backtesting and continuous drift monitoring, extended with bias and explainability controls for AI/ML, all on a tamper-evident audit trail.

The results

What changed

Illustrative, capability-framed outcomes from the modules deployed — representative of what the platform is built to deliver.

Every model now sits in one inventory with a current, evidenced validation status

Drift and actual-vs-expected monitoring flags degradation between formal reviews

AI/ML models carry bias, fairness and explainability evidence aligned to the EU AI Act and NIST AI RMF

Examiner questions on model lineage and validation are answered from one register, not reassembled by hand

We finally have one place that tells us which models are in use, when they were last validated, and which feed our regulatory numbers — without a spreadsheet hunt.

Head of Model RiskA top-10 North American P&C insurer

Modules used