SUCCESS STORY

Increasing Credit Card Approvals with Higher Accuracy.

Problem

A digital fintech in the Middle East provisioning credit cards was struggling to reverse rejection rates and was suffering from high first month default rates.

Credit Bureau information was limited for their target segment of blue-collar workers leading to low accuracy of their credit models.

Solution

We implemented our AI-powered credit modelling platform remotely:
Ensemble ML Application models
Bias reduction techniques implemented
Incorporated both existing customer data and Bureau data where available
Deployment completed in less than 4 weeks

Results

+20pts
Increase in GINI Accuracy
<1 week
to develop and deploy new credit models
+8%
Increase in Approvals
-5%
Gross Credit Losses