Case Study: UK High Street Bank achieves £750k annual savings and automates 57% of customer emails at 95% accuracy with Kortical AutoML

Customer Services Email Automation with AutoML

UK High Street Bank was struggling with a costly, slow back‑office process for small business account emails — over 500,000 incoming messages a year were manually read, triaged and copied into a PEGA ticketing system, causing backlogs and a minimum two‑day customer wait. The bank engaged Kortical, using the Kortical AutoML platform (Kortical AI as a Service) integrated with RPA to explore automated NLP triage.

Kortical delivered a production‑ready AutoML + RPA solution in four weeks that reads emails, routes confident cases into PEGA and sends uncertain items to human agents for review; human decisions are fed back to improve the model. The Kortical model now handles 57% of emails at 95% accuracy, continually learns from feedback, and has saved the bank over £750k per year.