Twenty-five years of enterprise AI, still happiest solving it with a team.

I'm an enterprise AI strategist, and I've spent the last twenty-five years leading AI and digital transformation programmes across Asia and the Middle East. Most of the time I'm brought in once a team already understands their problem and just needs someone who has shipped this kind of system before. That includes national smart city initiatives, critical infrastructure projects, and platforms at real scale, like KELIX, a private cloud RAG platform I built through IMDA's Digital Leaders Programme (2024 to 2026) that now serves more than 6,000 engineers with zero security incidents. I work the full deployment lifecycle: readiness assessment, governance design, production optimisation, ROI measurement, because I've seen what goes wrong when any one of those gets skipped.

My work leans on three things most AI implementations skip. First, structured impact assessment before deployment even begins, so the baselines and success criteria exist before anyone asks whether it worked. Second, AI governance that doesn't slow a team down: the oversight, escalation paths, and audit trails that let you move fast without losing the ability to explain a decision later. Third, a deployment approach that closes the gap between a promising pilot and a system people trust in production. It's the same Eight-Dimension framework I wrote up in my book AI at Scale: From Pilot to Production (forthcoming, Q4 2026), mostly so I'd stop re-explaining it from scratch on every engagement.

One framework I keep coming back to is Return on Employee (RoE). I introduced it because I got tired of AI business cases that only measured success as headcount removed. RoE looks instead at the capacity you get back per person: hours reclaimed from manual work, better decisions, more room to think. I've used it with mid-market business owners, government teams, and international development programmes alike, including at the UN ESCAP AI for Developing Countries Forum in Bangkok, and it tends to change the conversation in the room.

I'm finishing a DBA in Digital Leadership at Golden Gate University, and the research keeps circling back to the same test I already apply to client work: who benefits once a system is live, and can I stand behind the answer if someone pushes back. It's the same reasoning behind RoE and behind every governance decision above, not innovation as a talking point, but something a client, a regulator, or the person doing the job can trust.