AFTAs 2022: Best AI/machine learning initiative—Bank of America

Sherlock is a machine learning application linking incidents to change requests stored in the bank’s system of record. It was created to help the bank recover from business-impacting incidents faster.

Previously, incident management functions tended to be reactive and manual, and typically involved a triaged question-and-answer process. On average, it took roughly 90 minutes of manual effort to determine the root cause of change-related incidents