AI Readiness & Strategy
A clear-eyed read on where AI creates value in your business — and where it does not. I audit your processes, data, and existing pilots, then hand you a prioritised roadmap with effort and impact attached to every line.
Independent AI Enablement Consultancy
Most organisations do not have an AI problem. They have a dozen promising pilots, no route to production, and teams who have quietly gone back to the old way of working. I fix the route.
Engagements usually start with an assessment and grow from there. Each of these can stand alone, and often does.
A clear-eyed read on where AI creates value in your business — and where it does not. I audit your processes, data, and existing pilots, then hand you a prioritised roadmap with effort and impact attached to every line.
Hands-on programmes that take engineers, analysts, and product teams from cautious to fluent. Built around your codebase and your workflows, not a generic slide deck — people leave having shipped something real.
I build the thing. Agents, retrieval systems, internal tooling, and the evaluation harnesses that tell you whether any of it is working. Production-grade from the first commit, handed over with documentation your team can maintain.
The unglamorous work that decides whether the rest of it survives contact with the organisation: usage policy, guardrails, procurement support, and measurement that survives a board meeting.
Techstyles is Martyn Davies. It is an independent practice based in Košice, working with organisations across Europe, and it is one person: the person you meet in the first conversation is the person who does the work. There is no bench of juniors behind the pitch.
I have spent fourteen years on the adoption problem — the distance between a technology existing and people actually using it. I built and led developer relations and education teams at SendGrid, Algolia, Vonage and Orbit, and spent the last two years at Zuplo on how developers and AI coding agents adopt infrastructure. I am an Ambassador for the Agentic AI Foundation, the Linux Foundation initiative building open standards for agentic AI.
That history shapes how I operate. I would rather build a working prototype in week two than present a maturity matrix in month three, and every recommendation I make is one I would be willing to implement myself.
It also means I will tell you when AI is the wrong answer. A surprising share of the problems brought to me are better solved by fixing a data pipeline, retiring a process, or simply writing things down.
Building developer relations and education teams since 2012. The same person on the call, at the whiteboard, and on the pull request.
A short description of where you are is enough to start. I reply to everything within two working days, usually with a question or two.