Independent validation for AI and machine learning models
For financial institutions preparing for OSFI Guideline E-23 and the AMF guideline on the use of artificial intelligence.
Services
- E-23 and AMF readiness assessment
- For institutions whose model risk framework is not yet complete. A gap assessment against both guidelines and a prioritized remediation plan.
- Full validation of a high-risk model
- For models your framework rates high-risk. An independent review of conceptual soundness, data, performance, stability, fairness, explainability, implementation and monitoring.
- Targeted review of a lower-risk model
- For models rated low or medium risk. A shorter review of conceptual soundness, data and performance.
- Vendor pre-procurement assessment
- For institutions about to buy a model or an AI system. The vendor’s documentation, how the model behaves on your data, and the terms the contract should include.
- Generative AI pre-launch review
- For LLM-based assistants and document processing. Accuracy, bias, security, data provenance and human oversight.
Other services
- Revalidation
- Periodic reviews, and targeted reviews after a material change.
- Documentation dossier for model vendors
- Structured documentation and an independent review that a vendor can give to its institutional clients.
Independence
We don’t build the models we review, and our fees don’t depend on our conclusions. Our report is an independent opinion, and the decision to approve a model stays with your institution.
Background
PhD in experimental condensed matter physics (McGill University, 2019). Nearly eight years building and deploying models in industry: pricing algorithms at a proprietary trading hedge fund, search at one of the largest websites in the world, credit risk rating models at Shopify. Named inventor on four patent applications related to AI and machine learning. More about us