Product details

The case is set in January 2014 and reveals the data analytics initiatives at DBS Group Audit. Group Audit was used to assess the riskiness of the bank's branches based on seven attributes derived from the auditors' collective wisdom. The results could sometimes be misleading and inaccurate. To revamp this process, a machine-learning predictive modelling technique was introduced, and successfully correlated more than 130 risk-related attributes. In February 2014, DBS Group Audit and A*STAR's Institute for Infocomm Research (I2R) reached an agreement to set up a joint lab, leveraging the research institute's capabilities in developing innovative products and services. The outputs of data analytics are displayed in three forms. First, the heat map reveals the risk level of all branches in Singapore from a birds-eye perspective. Second, complaint analysis helps to identify customer needs more accurately, reducing the level of complaint. Third, cash discrepancy and headcount issues are reflected on the outcome page as a sudden jump in the graph. With these techniques, the auditors were expected to save man-hours and process the auditing work more effectively and efficiently by being responsive to changing risk profiles on a timely basis. The application of data analytics on risk profiling practices at DBS Group Audit is an anchor point for the bank’s vision of being predictive in risks. The bank is motivated to bring this data analytics initiative to other areas as well.