Learning from Change: Predictive Models for Incident Prevention in a Regulated IT Environment

ArXi:2604.13462v1 Announce Type: cross Effective IT change management is important for businesses that depend on software and services, particularly in highly regulated sectors such as finance, where operational reliability, auditability, and explainability are essential. A significant portion of IT incidents are caused by changes, making it important to identify high-risk changes before deployment. This study presents a predictive incident risk scoring approach at a large international bank.