Model risk governance is not the same as risk intelligence

Summary

Financial institutions have invested heavily in model governance frameworks, stress testing infrastructure, limit monitoring, data feeds, and dashboards, but risk leaders may still lack fast access to what those models are telling them. When a CRO must assess credit concentration, scenario sensitivity, or relationships between market positions and credit exposures, answering can require navigation across model outputs, analyst interpretation, and disconnected data systems. The proposed approach uses Databricks AI/BI Genie to let leaders query risk data in natural language, while retaining access controls, audit logging, Unity Catalog lineage, cross-risk data, and stress-test outputs in one environment. The stated distinction is that governance establishes necessary controls, whereas conversational risk intelligence supports questions that fixed dashboards did not anticipate, including comparisons with internal limits.

Context

Financial services institutions have sophisticated risk infrastructure, but CROs and other risk leaders may struggle to obtain fast, accurate, and defensible answers during credit committee reviews, market risk briefings, limit breach escalations, and regulatory inquiries. Pre-packaged reports answer anticipated questions, while relevant risk data and model outputs may remain separated across systems and interpretation layers.

Approach / What changed

Databricks AI/BI Genie provides natural-language access to a governed risk data environment. The described setup includes access controls, audit logging, Unity Catalog lineage, unified credit, market, operational, and liquidity risk data, and stress-test scenario outputs alongside actual exposure data.

Takeaways

• Genie queries and responses are logged, providing an audit trail intended to support model risk governance and regulatory examination readiness.
• Unity Catalog lineage allows Genie answers to be traced to the source data that generated them, supporting attribution requirements described for model risk governance.
• A unified environment combines cross-risk data with stress-test outputs, enabling questions about correlations, internal limits, and scenario performance beyond single-risk dashboards.

Related reading