Cloudera, the hybrid data company, today announced that ‘Buy Now, Pay Later’ (BNPL) leader, Afterpay, is using the Cloudera Data Platform (CDP) to run streaming analytics and advanced machine learning for real-time fraud detection. CDP helps Afterpay identify, access and reduce business risk relating to fraud for its 3.6 million customers, delivering on the company’s mission to power an economy which offers consumers better financial access, inclusion and experiences.

As Australia’s leading BNPL provider, the ability to ingest data from millions of transactions to identify duplicates and anomalies across its expansive customer base is critical. Given the eight-fold increase of its daily transactions from 100 million to 800 million in under a year, Afterpay required a more stable production environment to accommodate large-scale data to ensure analytics can be performed accurately at speed without compromising operational efficiencies. Existing in-house tools then were inadequate in managing these internal analytics requirements. The data platform also had to be flexible, scalable and built for growth to accommodate new and additional services and solutions to support Afterpay’s rapid expansion across the region and globally.

Harnessing the analytics capabilities of CDP Public Cloud, Afterpay’s engineers are now able to produce up to 800 million hybrid data streaming events and batch features for the data analysis team to perform fraud and predictive analytics for more data-driven and accurate insights.

Pinjie Huang, Engineering Manager, Afterpay comments, “Fraudsters today are becoming increasingly smarter in coming up with new patterns to take advantage of the system. We needed fast and flexible solutions to quickly adjust to newly identified fraud patterns or behaviours. The Cloudera Data Platform provided us with the right set of tools and visibility to proactively address these advanced fraud patterns by detecting issues and irregularities in real-time. As a result, operational efficiencies improved by 30% and internal net transaction loss (NTL) were kept under 10%.”

Afterpay further reinforced its fraud detection programme by adding Cloudera’s advanced machine learning capabilities onto its data platform, demonstrating CDP’s ability to scale. The data analysis team can now anticipate fraud activities by finding sequence patterns in valid transactions and flagging irregular transactions as potentially fraudulent. This resulted in a successful fraud rate reduction by 5%.

According to Keir Garrett, Vice President, Cloudera Australia and New Zealand, “Financial crime prevention requires continuous diligence and advanced analytics capabilities that accelerate detection and resolution. Given the surge in fraud activities, service providers need to up the ante in stemming BNPL fraud to give merchants and customers peace of mind over the security of their personal data. We are committed to supporting Afterpay in accelerating their data transformation journey and helping them unlock the power of machine learning and advanced analytics to manage business risks.”