Four engineers, twenty million dollars a month
A consumer lending platform that underwrites, prices and disburses a loan in about five minutes — and decides seven thousand applications a day without a human reading most of them.
Practice area
Lending & credit
Lending is the largest thing we do. Platforms that take an application from a phone, establish who somebody is, assemble enough evidence to make a credit decision, price it, fund it, and then service the loan for its life — in about five minutes, at a volume no underwriting team could staff.
7,000 applications decided daily
What this covers
Origination, decisioning and servicing. Machine-learned underwriting, alternate-data scoring for thin-file borrowers, and the collections and reporting behind them.
Case studies
—
Engineering notes from this work
07
Training a credit model on human decisions, not repayments
A new digital lender had no repayment history to train a credit model on, so we trained its approval models on about 8,000 decisions its credit officers had made by hand. The model inherits their biases too, and their own decisions are the control group for finding them.
06
Credit decisions from incomplete data: four models and an arbiter
The lending platform we built decides 7,000 applications a day, many from first-time borrowers with no credit history. Four separate models — credit bureau, bank statements, device, SMS — feed an arbiter that knows which evidence is missing and refers a case to a person rather than guessing.