United StatesFinance and insurance2025 – ongoingOppFi
Machine-learning underwriting model for loan pricing and automatic approval
OppFi approved 79% of loan applications without human review using its in-house underwriting models
Problem
OppFi lends to customers with limited credit history, so it has to judge risk and set prices across very different applicants. Manual review slows approvals and adds cost.
Approach
OppFi built its own machine-learning underwriting model, Model 6, to flag riskier borrowers and price risk across segments. Applications that clear the model are approved without human interaction. The company said it planned a refit, Model 6.1, for the first half of 2026. That refit would reweight variables using repayment data.
Results
Takeaway
A lender that refits its own model on repayment data can raise the share of applications it approves automatically. OppFi did not separate the model's effect from other growth drivers.
Categories
- Industry
- Finance and insurance
- Business function
- Finance, risk and legal
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Transcript of OppFi's fourth-quarter 2025 earnings call. The CEO said the auto-approval rate was 79% and that Underwriting Model 6 continued to bring benefits. The CFO called Model 6 a significant contributor to growth. A refit, Model 6.1, was due in the first half of 2026. The call did not quantify the model's own contribution.
Full details are in the original sources.