About

I have spent years inside organisations running agile transformations, and the pattern was always the same. Teams got their sprints, their stand ups, their certificates. Leadership kept making decisions exactly the way it always had, just with new vocabulary stuck on top. The problems teams raised in retrospectives, unclear priorities, unresolved dependencies, technical debt nobody wanted to own, got written down and left there, because fixing them needed authority that sat above the team. I'm watching the same thing happen with AI now. It's compressed into months instead of years, and it's considerably more expensive to get wrong. I started MML to do the part of this work most organisations skip. Not picking tools, not writing strategy, but working out whether the system underneath an AI investment can actually hold it: decision rights nobody's clarified, data fragmented across functions with no incentive to share it, governance that exists on paper and nowhere else, leadership announcing change without doing anything to make it land. That diagnosis is the work. A transformation only succeeds if it's built on an accurate one.

I started my career at GCHQ, where security and governance came first rather than getting bolted on afterwards. That ordering has shaped how I look at every system since. My background is agile coaching and delivery, at enterprise level at Jaguar Land Rover, at Dyson, and most recently at Entain, on their own AI adoption in real time. That kept the diagnosis grounded in what's actually happening on the ground, not theory.

About

I spent years watching transformations fail for the same reason.