I lead whole-system AI transformation and coach the leaders carrying it.

Screen, interview, try me on real work, then commit only while it earns its keep. You can stop at any step, and each paid one ends in a written artefact you keep. Start wherever the risk feels right for you.

The rest of this page is the thinking behind it, then the detail on each step.

If what you need is an operating role, fractional or permanent, I’m open to that too. Say so on the call.

Your AI spend is up and your cycle times haven’t moved. We start there—in your queues, your cost-of-delay, where decisions actually route—not in a framework. I carry one prior and I’ll name it so you can hold me to it: the gains tend to land inside a delivery structure designed before these tools existed. We read your numbers together, and they settle it. They show the constraint, price it, or kill the prior, and if they kill it we stop. Nothing gets installed on a hunch, mine or yours.

Before I start, you do the staging: line up a handful of interviews across one value stream (an hour each, including the executive who owns it and someone close to the work), send out a short story-prompt survey I’ll hand you, and get me access to your work-tracking tool or a dump of its data. Then we spend up to two full days on it together: I read your queues and your people’s own accounts, run the interviews, and we look at what surfaces side by side. The written decision note at the end says what looks constrained, what’s still unknown, and whether the next step is stop, periodic advisory, or the diagnostic. This is your situation worked on directly, not a workshop with seats; I don’t sell training days, and this doesn’t become one. Already run your own numbers? Bring them: we build on your draft.

The clock starts at kickoff, after a short mobilisation gate: interviewees named, the finance session booked, data access agreed, so neither of us pays for scheduling drift. There are no automatic rollovers and no bundled follow-on. We continue only if the experiment’s numbers argue for it.

Your sponsor (usually the CTO or VPE), one director, the product counterpart, one engineering manager, and someone from finance who can price delay. The CEO joins twice, at kickoff and at the decision review. The two touchpoints are deliberate: the diagnostic also tests whether the CEO engages with the diagnosis, because everything downstream depends on that.

Built with your people on the value stream you choose, and shaped to what it needs. A typical engagement produces something like:

The embedded work is hands-on: four days a week, and I roll out the redesign myself—a LeSS adoption where it fits, a bespoke SDLC where it doesn’t—the way I spent 15 months re-shaping a global firm’s operating model as the sole practitioner across 20+ teams, with tooling I built carrying the mentoring load and the CTO holding the authority the design needed. I will propose it only if three things are true at the end of the diagnostic:

After the trial the engagement simply continues on the same flat fee until either of us ends it on 30 days’ written notice.

Training days by the dozen, certification programmes, a coach army, or transformation on a timeline. My coaching isn’t a workstream running alongside the redesign; it’s how the leaders carry the change. It commits to process only (cadence, supervision, review); the outcomes belong to the redesign itself, where they can be measured. How that coaching runs—method, confidentiality rules, limits—is public at /coaching-practice.

From leaders I’ve done this work with (excerpts from LinkedIn recommendations):

Working from New Zealand; my standard delivery shape overlaps US Pacific business hours Monday to Friday (Here’s the overlap visualised).