Models · Credit / Risk

Credit Oracle

Federated credit risk model for SME lending, trained on anonymised loan tape from 14 partner lenders.

  • Risk: medium
  • Beta
  • Federated
  • US
  • UK
  • EU
AUC (out-of-sample)
0.84
Gini
0.68
PSI (vs baseline)
0.04
Federated lenders
14

What it does

Returns a probability-of-default and a recommended pricing tier for an SME credit application. Trained federation-style across 14 partner lenders without any borrower-level data ever leaving its lender of origin.

Why it works

Federated training pools the statistical signal across lenders without pooling their data. The result is a model with materially better out-of-sample AUC than any single lender’s bespoke model — without raising the data-privacy and competition-law issues that block centralised pooling.

Compliance posture

Each prediction comes with a SHAP-style explanation suitable for adverse action notices in the US and the UK. EU deployments include the additional documentation required under the EU AI Act for high-risk credit-scoring systems.

Pricing

Pro plan and above. Federated lender onboarding includes a data audit and an explainability review.

Audit & jurisdiction

  • US — reviewed and routed to local counsel.
  • UK — reviewed and routed to local counsel.
  • EU — reviewed and routed to local counsel.

Every run of this model is appended to a hash-chained audit log. Anyone with a run id can fetch a Merkle inclusion proof — see the Compliance & Trust page.

Reported metrics are self-attested by the model author and verifiable against the audit log.