Case Study: BGL BNP Paribas achieves rapid, accurate fraud detection and enterprise-wide data democratization with Dataiku

Improving Fraud Detection by Evangelizing Data Science

BGL BNP Paribas, one of Luxembourg’s largest banks and part of the BNP Paribas Group, struggled with a largely static machine-learning fraud model because of limited data science resources and restricted access for business teams. To democratize data access and modernize anomaly and fraud detection, BGL BNP Paribas partnered with Dataiku and implemented Dataiku Data Science Studio (DSS).

Using Dataiku DSS, cross-functional teams built a new fraud detection prototype in eight weeks and leveraged Dataiku’s enterprise security and production features to move quickly into production and show results soon after project start. The work enabled rapid sandbox prototyping (new use cases tested in a few weeks), preserved governance, and catalyzed a cultural shift toward deployment and industrialization—BGL BNP Paribas has already launched three additional data projects following the initial success with Dataiku.

BGL BNP Paribas