Case Study: HSBC achieves real-time analytics (6h→6s), a single Delta Lake, and 4.5× PayMe engagement with Databricks

Reinventing mobile banking with ML

HSBC, a global bank with 39+ million customers, developed PayMe, a social mobile payments app that quickly grew to 18+ million users in Hong Kong. The team struggled with legacy systems and slow, manual data pipelines—analytics were taking hours and data was often stale—while data masking, siloed data science, and 14 separate read-replica databases limited their ability to deliver personalization, fraud detection, and real-time insights.

By adopting Azure Databricks and Delta Lake to unify data engineering, analytics, and ML, HSBC implemented real-time anonymized data access, scalable pipelines, and collaborative model development. The change cut complex analytics from 6 hours to 6 seconds, consolidated 14 databases into one Delta Lake, enabled predictive use cases (recommendations, anomaly/fraud detection, transaction enrichment), and helped PayMe capture 60% market share in Hong Kong with a 4.5x improvement in app engagement.

HSBC

Alessio Basso

Chief Architect