The short version
I started in web engineering and stayed for the results. Early work was WordPress sites that were slow to load and impossible to rank; the fix was not a plugin, it was learning how the stack actually behaves. Today I build on custom, high-speed stacks — static and server-rendered where it fits — so a page is fast before anything else is said about it.
That same discipline runs through the search work. Technical SEO is not a checklist to tick; it is Core Web Vitals, schema markup, internal architecture and multilingual structure, engineered properly by people who do it every day. Paid acquisition sits on top of the same foundation: campaigns built on intent, landing pages built to convert, and a number attached to every click.
The last piece is automation. A site that generates enquiries is only half a system; the other half is what happens next. We wire the CRM, the booking pipeline and the follow-up so nothing depends on someone remembering to check an inbox.
I do not do all of this alone. I lead a focused execution team — design, engineering, content, media buying and automation — and I stay the lead strategist on every account. You get one person who is accountable for the whole thing, not five vendors pointing at each other.
Who I work with
Medical clinics and dental practices that need more booked appointments from the same advertising budget. Small and medium businesses whose website is not earning its keep. Exporters and B2B sellers who need to be found in markets that do not speak their language. And founders who want to build a lean AI or SaaS product with a technical partner instead of a freelancer.
How I think about the work
Coordination cost, not capability, is what kills agent systems and marketing programmes alike. Most "AI transformation" fails at the process, not the model, and automating a broken process only makes it break faster.
So we build the same way each time: an input that gets validated, a step that is allowed to fail loudly rather than quietly, and an output a human can check cheaply.
What I won't do
I won't state a number I cannot link to. Where a figure is unknown, I say it is unknown. I won't sell certainty about the AI roadmap, because anyone doing that is selling something other than a forecast. And I won't run fake urgency: no countdown timers, no seats that quietly reappear.
Client names, revenue figures and internal details stay out unless I have permission.