Case Study: Leading Global Payment Solutions Provider achieves 20% transaction fraud detection uplift and 5x faster model builds with DataVisor

A Top Payment Network Relies on DataVisor’s AI Solutions to Boost Transaction Fraud Detection

Leading Global Payment Solutions Provider, which processes more than $8 trillion in payments annually, faced rapidly evolving fraud tactics, fast model decay, long model-build times, incomplete digital data and an aging infrastructure that could not support real-time, cross-channel detection. To address these challenges the company engaged DataVisor, using DataVisor’s Enterprise ML platform and unsupervised machine learning to augment its existing rules-based and supervised models.

DataVisor implemented a combined solution—unsupervised + supervised learning, a feature-engineering layer that works with imperfect data, and a high-throughput big-data platform for real-time scoring—to detect correlated and emerging fraud patterns. As a result, DataVisor helped the client achieve a 20% uplift in transaction fraud detection, 94% detection accuracy and a 5x faster model development cycle (from 4–6 months to weeks), while enabling low-latency, production-ready fraud detection at scale.