Model Risk, Reimagined: Why GenAI Is Reshaping Governance, Not Just Math

In this insightful interview recorded at CeFPro’s Risk Evolve convention, Alberto Ramirez, Partner at Solytics Partners, outlines how generative AI is challenging traditional boundaries of model risk management (MRM). Ramirez emphasizes that model oversight now extends beyond risk and validation teams to include compliance, legal, and technology leaders—a necessary shift given the opacity and complexity of large language models.

To meet this new reality, Solytics Partners has developed a modular ecosystem of tools: Nimbus Uno for automated MLOps and documentation, MoDeva for advanced statistical validation, and MRM Vault for governance and traceability. These tools work in unison to embed model controls deep into operational workflows, supporting transparency across the lifecycle—from development to deployment.

Ramirez stresses that regulation, often seen as a burden, is in fact a catalyst for smarter tech development. By designing configurable systems, Solitix ensures institutions can adapt swiftly to evolving regulatory demands. Rather than taking a reactive stance, firms should see automation as a proactive way to embed compliance guardrails into scalable infrastructure.

Looking ahead, the future of model risk lies in cross-functional integration and renewed clarity on material business impacts. Ramirez argues for a return to basics: identifying what really breaks a bank—not just managing models for the sake of models. Events like Risk Evolve, he says, are key to staying grounded in the industry's real-world challenges.