Predict Default Rate

A micro-lending company in India built a model to identify customers who are likely to default, automated decision making and increased repayment.

Fintech
Analytics & BI
100,000

rows of data

5 days

time to value

79.2%

prediction accuracy

7%

decreased defaults

Background

“Obviously AI is a no-brainer tool for predictive analytics. Easy to integrate and predict outcomes without any knowledge of machine learning.”

- Namra Parikh, COO at Creditt India

Obviously AI can be used to estimate default probability, loss severity, and for loss forecasting, using past client behavior data. These predictions improve pricing for risk, credit approval, and portfolio management. Obviously AI also automatically updates, making credit scoring more precise as models learn the nuances of discrete populations.

Creditt is a Micro-loan providing platform that is affiliated with some of the major banks in India - Kotak Mahindra Bank, ICICI Bank and many more. Creditt loans out between $10,000 and $50,000 per month to their 750,000 customers.

To provide these loans, Creditt would have employees manually read applications and accept/reject them. This meant tiny details slipping through the cracks and several customers ending up defaulting. So they decided to turn to Obviously AI instead.

The Solution

Creditt conducted extensive market research and chose Obviously AI’s predictive analytics platform for their needs. The Analytics team used Obviously AI’s platform to train a model to analyze their customer’s demographic and behavioral data to answer the following questions:

1. How much money to loan this customer?
2. Will this customer pay money back on time?


They used Obviously AI’s REST API to integrate predictions directly into their app directly displaying the numbers to customers.

The Results

Creditt’s customer default rate decreased by 7%.

Obviously AI’s platform allowed Creditt’s existing analytics team to build the AI model within 3 days, versus 4 months it would require a data scientist, and at a significantly lower cost.

Most notably, Creditt empowered its existing analytics team with the capability of building state of the art AI predictive analytics models, without the need for an on-call data scientist.

See how no-code machine learning can transform your business. Book a demo now!

Get Started Now

See how no code machine learning can transform your business and change how you make decisions.