Personal Loan Score
A personal loan score trained on your own borrowers and loan performance.
The Personal Loan Score predicts whether a borrower will repay, using their bank transactions alongside application and bureau data. Use it to approve good borrowers your current model declines, and to set rates that match each borrower’s real risk.
Built on millions of anonymized personal loans across multiple lenders. No PII required.
Trusted by
Every personal lender has near-miss applicants its model can’t sort with confidence. Some of them would repay.
Approving all of them isn’t the answer and that’s where the Personal Loan Score comes in.
Approve more with a second look
Some applicants you decline today have steady income and enough left each month to cover the payment. Pave measures income stability, cash reserves, balance trends, and spending relative to income from bank transactions, so you can approve the ones who can afford the loan.
Set APR and amount by predicted repayment
The Personal Loan Score is trained on anonymized repayment outcomes from millions of personal loans across multiple lenders. Use it to set the APR and loan amount for each applicant.
Set payments to payday
If a payment is due on the 1st and the borrower is paid on the 3rd, the payment can bounce even when they can afford it. Pave finds each borrower’s pay date from their bank transactions and sets the due date just after it.
Customer story
How a national lender underwrites with a custom Personal Loan Score
A national lender worked with Pave to build a score trained on its own loan history. The score adds bank transaction data to the lender’s existing underwriting.
Trained on the lender’s loan performance history and consumer-permissioned bank data.
The score learns which transaction patterns came before repayment and default in this lender’s book.
Pave and the lender’s risk team track the score’s performance and retrain it as the portfolio changes.
How it works
A Personal Loan Score built on your loans
Pave trains the score on your borrowers and their repayment outcomes. Your team can review every feature that goes into it. No PII required.
Your funded loans and repayment or default results become the training data.
It starts from a score trained on millions of personal loans, then retrains on your own loans and the outcome you care about.
We show the approval, loss, and pricing impact before you go live.
We track the model’s performance after launch and retrain it when your portfolio changes.
See it on your own book.
Send us past applications and their outcomes. We’ll score them and show how your approvals, losses, and pricing would have changed.