Specialist Scale
My productised agency: patient acquisition for self-pay private medical practices, owned and run end to end.
Work with me
Sitting on a product or AI decision you can't afford to get wrong?
That is the conversation I want to have.
You already know the decision. The room is busy, high energy, and not yet agreed. I sit with your team until the next quarter is obvious, and I stay while you build it. I do not build it for you. Most engagements start small and earn the next step.
How we work
Most engagements start small and earn the next step. Start light, step up without switching.
Workshop
A working session with your leadership and team to get aligned on the product's biggest levers and where AI earns its keep. The lightest way to start.
Half day
Product and AI Strategy Sprint
A sprint run with your leadership, product and engineering, not handed over as a report. Week one I dig in: interviews, your product, roadmap, data and stack. Week two we work it through together: the few levers that actually move the product, where AI belongs and where it does not, what to fund and what to kill, and the sequence to bring it in. You leave with one decision-ready strategy memo, an orchestration plan for who builds what and in what order, and a team that owns it. The fee credits toward a fractional engagement if we continue.
Two-week window
Fractional Product Advisor
The spine. Ongoing product judgment, strategy, data, platform and where AI fits, plus coaching your team and orchestrating the build alongside them. Two intensities, so you start light and step up without switching.
- Light. A few hours a month, a sparring partner for a founder or product lead.
- Core. One to two days a week, embedded, owning the thinking and the operating discipline. Three-month initial term, then rolling with 30 days' notice.
Ongoing
Pricing is shared on the fit call, once we both know the shape of the work.
Track record
Twenty-two years building product across Sun Microsystems, EMC², Dell, Microsoft, Skype, Tesco, Hiya and Smartnumbers. At Hiya the useful move was not a bigger model. It was a honeypot network: phone lines that only scammers call, so the system could learn. That is how you reach billions of calls a month without guessing. I bring that kind of judgment independently now.
Proof · I run this myself
Most advisors have never carried the risk. I prove the method on my own, in one vertical, before it reaches a client.
Bring me the decision you are sitting on. Let us get it right together.
A 30-minute call to see if it is a fit. No pitch.