I act as your owner’s representative on agentic AI: I establish whether the move is worth it and whether to build or buy, write the spec, select the implementation partner and govern the work — EU-compliant from day one. I never bid on the build. That is not a question of capacity; it is what makes the sign-off worth anything.
Whether agentic AI is worth it at all, and then build vs buy. Clear stage-gates, not another pilot graveyard.
I write the spec, run the selection and hold the partner to the stage-gates. Because I never bid on the build, the stage-gates have teeth.
AI Act, CRA, NIS2 as a design input from day one, not a blocker at the end.
Written into the contract. And the other half of it: no company I advise on go-to-market goes onto a partner longlist. Where that overlaps, I disclose it and step out of the selection. Backed by eight years in standards governance: four on the OSGi Alliance board, four on the Steering Committee of the OSGi Working Group at the Eclipse Foundation, co-chair of the OTA SIG at the Industrial Internet Consortium.
You end with an answer on whether agentic AI is worth it for your case — including a no. In writing, with reasoning and a cost envelope.
Build, buy or platform: compared on cost, dependencies, and what you will still be able to change in three years.
Longlist, selection criteria, negotiation support — and governance of the chosen partner through to sign-off against the spec.
AI Act, CRA and NIS2 shape the architecture from day one instead of surfacing as a blocker at the end — including who carries which obligation.
Defined stop and go criteria per stage. A pilot that does not hold up gets ended, not extended.
Ownership, operating model and governance end up with named people on your side — not in a document.
"Implementing Agentic AI: Building Your Organizational Playbook", MIT Sloan Executive Education with the MIT Schwarzman College of Computing.
Responsible for the Hybrid-AI showcase with a documented first-pass-yield improvement and a 90% reduction in report preparation time.
A full app built solo with Claude Code, agentic architectures and prompt engineering done myself. Anything I put in front of a CTO I have run on my own keyboard first.
Architecture review, roadmap and AI readiness in four weeks. Written memo with current state and a 12-month roadmap, kickoff and closing workshop.
Industrial AI roadmap and the translation of concrete customer problems into engineering and model requirements. Ends with a roadmap the board and engineering both own.
Ongoing support on roadmap, governance and partner management — quarterly or as needed, without a permanent hire.
No, deliberately not. I ran a delivery and GTM organisation of some 40 people for ten years and carried FOTA at OEM scale — I could. That is precisely why I do not bid: whoever picks the partner and signs the acceptance must not earn from the contract. No vendor I recommend pays me anything, and that is written into the contract.
From recommendations, never. No vendor I put in front of you pays me anything, in any form. The complete answer needs the other half of my work, though: I advise industrial-tech scale-ups on go-to-market, sometimes with a success-based component or advisor equity. So the rule is hard: no company that pays me that way, or that I hold a stake in, enters a selection process I am running. Where that overlaps, you hear it from me before it matters, and I step out of the selection.
Then that is what the memo says. It is one of the more common outcomes, and it usually saves considerably more than the audit cost.
The maturity check takes five minutes and costs nothing. An audit delivers a written result after four weeks. First verifiable results inside a mandate: within 60 days.
Directly with me. No associates, no junior layer. Where specialised depth is needed I draw on a small trusted network of senior freelancers — by exception, with clear rationale.
Seven questions, five minutes. You get your stage, the typical trap, and the highest-leverage next move. No email required.