Case Study: a debt collection agency improves collection effectiveness with InData Labs predictive analytics

a debt collection agency boosts collection efficiency 2x with InData Labs

The customer, a debt collection agency, sought to improve the effectiveness of its debt collection process with predictive analytics. They partnered with InData Labs to develop a machine learning model that could identify customers most likely to repay their debts.

InData Labs built and deployed a predictive model using Python and LightGBM on the client's existing MS SQL Server infrastructure. The solution accurately forecasts the probability of a debtor's promise to pay, achieving a significant ROC_AUC score of approximately 0.775. This allowed the agency to optimize its agents' time by prioritizing the most promising accounts, resulting in a two-fold increase in revenue.