Situation
The decision speed and quality of credit application assessments are vital for the success of financial institutions. Manual assessment of those applications is a slow and expensive process that requires reviewing multiple information sources, and inconsistencies may not always be noticed.
Approach
By using advanced machine learning techniques, Valcon created a model that combines application data with information from credit risk agencies to optimise decisions. The model classifies credit applications and provides advice on acceptance or rejection.
Results
The model that was developed for our client increased the number of applications that could be handled automatically by 12% (from 70% by classical modelling to 82% by advanced machine learning). The addition of extra data only improved these results. A second model for international markets is currently in development.
