AgentHive/case-studies/gig-economy-credit
All case studies
Gig Economy

31% reduction in income-uncertainty defaults for gig workers

An online lender serving independent contractors replaced static income underwriting with a cash-flow volatility model — cutting defaults without shrinking the book.

−31%default rate reduction

Context

A fintech lender discovered that gig-economy borrowers had 2.4× the default rate of traditional W-2 borrowers at equivalent stated DTI ratios. Volatility in monthly income — not income level — was the root cause.

What we built

AgentHive's feature engineering pipeline constructed a 12-month rolling income-volatility index from bank-feed transactions. The rules engine added an income-stability gate on top of the ML score, with automatic SHAP reason codes for every decline.

Results

MetricBeforeAfter
12-month default (gig cohort)8.9%6.1%
Approval rate (gig cohort)54%57%
Manual-review queue depth38% of pipeline7% of pipeline
Regulatory exam findings20
"We didn't shrink our book — we just stopped lending to volatility risk we couldn't see. AgentHive made that volatility visible." — Head of Risk

See it live in the Interactive Lab

Run the same decision scenarios yourself — no setup required.

Open the Lab