Upstart, an AI lending platform, today announced its Credit Decision API. By tapping into Upstart's AI models, banks and other lenders can deliver instant credit decisions for auto loans, personal loans, and student loans with higher approval rates and lower loss rates. The service's programmable interface means lenders can integrate AI-powered decisioning into their existing infrastructure and workflows.
"Upstart's AI models have significantly outperformed traditional FICO-based models before and during the COVID-19 crisis," said Dave Girouard, co-founder and CEO of Upstart, "By leveraging AI, banks can improve and help de-risk their lending activities during even the most rapidly changing economic environments."
Incorporating more than 1,500 variables, Credit Decision API leverages machine learning algorithms trained on more than 550,000 loans and more than 7,000,000 repayment events across Upstart's network of bank partners. Compared to traditional bank models, AI-enabled decisioning can provide banks with up to 75% fewer defaults at the same approval rate.*
Upstart's Credit Decision API is designed to provide banks with complete authority and control over their lending programs. Each decision is based on the bank partner's credit policy, lending terms, and return targets, in addition to the risk modeled for each applicant.
Upstart's Credit Decision API integrates seamlessly with any digital lending platform, enabling risk-based decisioning and declines (including Adverse Action Notice creation). Banks use their existing digital flow to gather just name, address and income information, which are passed via secure API. The results are displayed within the bank's existing application workflow.
Features of the Upstart Credit Decision API include:
- Enables instant credit decisions with just name, address, and income provided by applicants
- Expanded data support available for enhanced model accuracy
- Decisioning assistance for auto loans, personal loans, and student loans
- Dynamic response of loss rate modeling to changing unemployment levels
- Programmatic enforcement of lender's credit policy, fees, and credit terms
- Automated generation of intuitive Adverse Action Notice for all declines
- Customizable return curve to match lender's yield targets as a function of risk
Additionally, the API can also be used to evaluate existing pools of loans in order to estimate future loss rates, helping banks to refine their loan loss reserves and servicing plans.