Obvious train wreck was obvious.
What was more likely, CX to use CMP project to take a fundamental review of rostering methods, then implement a new technical solution around that, or to take what they have, and insist on trying to automate it "better", then blame the system when that doesn't work.