Case Study: Coins.ph achieves 70% reduction in operating costs and faster ML-powered fraud detection with Databricks

Reinventing digital payments using data and ML

Coins.ph, a leading digital payments platform in the Philippines serving over 10 million customers, needed to perform accurate real-time financial audits and prevent fraud and money laundering. Their legacy AWS/EMR analytics stack was costly to maintain and slowed ML experimentation, hindering the company’s ability to deliver timely, data-driven features for remittances, bill payments and e-commerce.

By adopting Databricks (Delta Lake, MLflow, interactive notebooks and automated cluster management), Coins.ph unified data and ML workflows, enabled near-real-time analytics, and sped up model development and cross-team collaboration. The move cut infrastructure costs by about 50% and operational costs by about 70%, reduced time-to-market for ML features from weeks to days or hours, and strengthened fraud detection, AML compliance and recommendation capabilities.

Coins.ph

Dmitry Ustimov

Data Architect