Case Study: Leading International Bank achieves 15x reduction in false positives with Bottomline's CFRM platform

International Bank Cashes in on Fraud Investigation with Intelligent Machine Learning

An established international bank experiencing a steady rise in payment-fraud alerts engaged Bottomline to address the problem. By 2016 the bank processed about 250,000 transactions daily and was seeing ~1.5% of transactions flagged (~3,750 alerts/day) even though only 8% were true fraud, forcing investigators to spend roughly 20 minutes per alert and stretching resources.

Bottomline upgraded the bank to its CFRM platform, which uses intelligent machine learning, rules-based detection and behavior profiling to score and triage alerts into actions (Stop Transaction, Call Customer, Unsuccessful Call). The prioritized “fraud likelihood” scoring reduced investigative time (saving ~10 minutes per call), let the same team handle more alerts without extra headcount, and delivered a 15x improvement in reducing false positives.