What I believe about AI, leadership, and building systems that work when it matters.
The defining challenge of enterprise AI is not capability. It is the distance between what works in a demo and what works when a regulator, a board, or a real user is watching. Most organizations are stuck on the wrong side of that gap — running pilots that never reach production, building agents that hallucinate in ways no one catches until it’s too late, treating governance as a cost center rather than the infrastructure that makes scale possible.
I believe the organizations that win in AI will not be the ones with the most models or the biggest compute budgets. They will be the ones that treat AI as an organizational capability — something you build teams around, not just tools for. That means hiring leaders who understand both the code and the consequence. It means building evaluation frameworks before you build agents. It means treating production readiness as a first-class design constraint, not an afterthought.
I have spent the last decade working at this intersection — building production AI systems for clients who cannot afford a wrong answer, and building the organizations that sustain those systems over time. The lesson I keep learning: the technical problem is rarely the hard part. The hard part is context management — making sure every person and every system has the right information at the right time to make the right decision.
This is true for distributed engineering teams. It is true for AI agents. And it is the lens through which I think about everything I build.