Prove launched Signals, a new POST /identity/intelligence/v1 endpoint that takes a name and phone number and runs Prove's identity intelligence model on the identity behind the number. The response returns a typology (such as Verified, Household Verified, Phone Risk or Unverified), a Low, Moderate or High risk level, at most one indicator code explaining the result (such as LOW_TENURE), and the list of feature values the model used, including name match scores, identity tenure, phone query activity and IP details. It is US only and testable in Sandbox.
Our readBanks and lenders get a model output that carries its own reason code and the input features behind it, which supports explainability and model validation work when using the result to drive approve, step up or decline policy.