top of page

This auto loaning company required a model to predict defaulters through a machine learning system


USD saved each year

The estimated positive impact each year the ML system we created was over USD 600,000


Revenue increase

The ML was able to accurately predict possible loan defaulters, thereby increasing revenue


Effort reduction

The effort, time and money required to pursue loan defaulters dramatically reduced

The auto loan company based out of the US possessed revenues exceeding USD 140 Million. However, they had to work with a department that tried to call up loan defaulters and even having to visit them to get them to pay up. Seldom did they succeed

The client is a $20 billion IT giant with operations across the world providing mission-critical IT services. With operations in 70 countries globally, the client drives innovation in the IT world. The client has over 130,000 employees across the world and is a Fortune 500 global IT services leader.


Having understood the challenge, we worked with the internal teams to see if they could identify a pattern among loan defaulters. Using advanced machine learning systems, and performing vigorous data, analytics and insights, we built a model that could predict customers who are likely to default


The key challenge in this case was the complexity to get the defaulting loan

  • The client could not identify which customers are likely to default on loans

  • With increasing defaulter, there had to be increasing operations

  • Revenue was poured into pursuing defaulters through various means


Through machine learning, we were able to

  • Build a model that was able to predict possible loan defaulters

  • The ML system made weekly predictions and fine-tuned itself

  • Data we assessed included loan history, payment history, emails and calls

Download the case study to know more about the benefits we delivered, and how we executed this project





Incident Response

Lorem ipsum dolor sit amet, consectetur adipiscing elit.  consectetur adipiscing elit. 


bottom of page