Metro Bank cuts transaction monitoring alerts 20% with SymphonyAI financial crime platform

Metro Bank's existing financial crime systems, first deployed nearly a decade earlier, were no longer aligned with the demands of a fast-moving digital banking landscape. The legacy platform could no longer meet operational needs, and enhancements to the software were difficult to introduce without significant overheads. The bank needed a unified, cloud-native, future-ready platform that could consolidate the full financial crime lifecycle, support dynamic customer risk assessment and enable continuous innovation without disruptive overhaul.

Metro Bank selected SymphonyAI's platform as the most comprehensive and cohesive solution and the best foundation for AI-driven capabilities. The new SaaS suite consolidated transaction monitoring, payment screening, name screening, AML investigations, customer risk assessment, legal orders, court orders and ongoing due diligence in one integrated ecosystem, with customer risk assessment embedded in the same platform as monitoring and screening. Machine learning models and intelligent exclusion rules that the bank had previously implemented outside its legacy solution could now reside inside the platform, aligned with model governance. The deployment ran in two tranches, with payment screening completed in approximately 4 hours and the larger second tranche completed across a weekend within 24 hours.

Under the new architecture, alerts are rolled up to the customer level, reducing unnecessary duplication and producing a measurable 20% drop in transaction monitoring alerts shortly after go-live. One-call customer onboarding now replaces two separate processes, which reduces unnecessary reviews. The bank can also implement updates in an agile but controlled manner, and the platform includes built-in data governance, GDPR-aligned controls and embedded model governance.

• Consolidating monitoring, screening and customer risk assessment on a unified platform lets a bank act on internal and external risk triggers dynamically.
• Rolling alerts up to the customer level reduces duplicated alerts.
• Bringing existing machine learning models inside the platform places them under model governance.