Skip to main content
BDI

Risk & compliance

Model Validation

Independent review of a model's design, data, implementation, performance, and limits.

Buying context

Why it matters

Models influence estimates and decisions but can fail through bad assumptions, code, data, or use. Validation should explain where a model is reliable, where it is not, and how findings affect use.

Due diligence

Questions to ask

  1. How will the validator establish independence from the model developer and business owner?
  2. Which conceptual, data, implementation, outcome, and governance tests are in scope?
  3. How are vendor-model limitations evaluated when source code or training data are unavailable?
  4. Will findings state the affected decision, materiality, compensating control, and required follow-up?

Alphabetical, not ranked

Company listings

The directory does not yet list vendors in this category. Companies can review the listing terms and contact Bank Data Insights to be considered.

Related categories