Comparison
Keep it while you test. A Carrington Labs model is built around data you want to use, including attributes from bureaus and bank transaction data. Want to test how a custom model compares to the standard attributes you're currently using?
The practical difference is control. Carrington Labs can combine cash flow and traditional data, engineer features for the lender's population, calibrate the model to the lender's outcomes and align the strategy to a defined commercial objective.
FICO scores provide a standardized measure of consumer credit risk. UltraFICO adds customer-permissioned cash flow data from Plaid to FICO credit information.
Carrington Labs can use cash flow, bureau, application, product and internal performance data together in one lender-specific model. The model is built against the lender's own observed outcomes and can be aligned to the result the lender is trying to improve, such as stronger risk separation, lower delinquency, controlled approval growth or improved margin.
Carrington Labs develops the target, features, calibration and validation around the lender's own product and repayment outcomes. A standardized score is not built around the lender's specific product structure, customer mix, acquisition channels, policy and loss definition.
Carrington Labs supports cash flow underwriting without forcing a lender to choose between transaction data and traditional credit data. The model can use the combination that is most predictive for the lender's portfolio.
Carrington Labs converts raw transaction, bureau, application, product and repayment data into predictive lender-specific features. As additional performance history becomes available, retraining is included and new features can be engineered where the expanded data supports them.
FICO and UltraFICO provide a credit score. Carrington Labs can also use the resulting probability of default within the Credit Offer Engine to recommend amount, term and price using expected loss, revenue, funding and direct costs, expected take rate, price elasticity and lender constraints.
Carrington Labs is built for lenders that want:
For custom model engagements, proprietary lender data remains within that lender's model-development and validation process and is not pooled into another lender's custom model, subject to the governing agreement.
Use your historical portfolio to test whether lender-specific feature engineering and calibration improve risk separation and expected commercial value.
Comparison based on publicly available product information as of July 2026. Third-party names and trademarks belong to their respective owners. Carrington Labs is not affiliated with or endorsed by the companies referenced unless expressly stated.
Leave your details below and our team will schedule a call to discuss your business and how Carrington Labs can help.
