On the long game.

I'm a principal-level software engineer and architect — an individual contributor by deliberate choice, more than thirty years in. I've designed and delivered enterprise software for federal, healthcare, financial-services, utilities, telecommunications, transportation, education-assessment, and commercial environments — the kind that runs quietly in the background of consequential work and has to keep running long after launch. I came to it through engineering before computer science, which is probably why I still think about software the way I think about machines: as a real thing with constraints, failure modes, and a tendency to surprise you if you stop paying attention.

What I've actually learned, over and over, is that the technology is rarely the hard part. The hard part is fitting the work to the people, the constraints, and the timeline — and not pretending the surprises won't come. So I've made it my job to be the steady hand on a build: clear about the architecture, honest about the tradeoffs, and the person a team can look to when the problem is hard and the stakes are real. That trust is the thing I'm proudest of, and it was earned one delivery at a time.

In recent years my work has moved into applied AI. I build production-grade language-model systems — retrieval pipelines, agentic workflows, the evaluation and observability that keep them honest. The principles didn't change; the material did. Most of the difficulty still isn't the model. It's the engineering around it: making the thing legible, measurable, and trustworthy enough to put in front of people who are counting on it. And the line I hold is simple: AI where reasoning or language genuinely helps; deterministic, testable code wherever money moves, access is granted, or a number has to be right.

The way I produce work has changed too, deliberately. For most of my career the craft itself was the point — knowing the language cold, typing every line, taking pride in the how. The era we're in rewards a different posture: real clarity about what is worth building, and the discipline to get there with AI coding agents doing more of the typing. I decide what to build and how it should hold together; the agents help me move. The standard underneath is the same one I've always held — build it so it works, prove it with tests, and leave it clean for the next person.

None of this is the whole of me. I'm a family man and a man of faith, and that's the quiet center the rest turns around — it keeps me patient, grateful, and clear about what actually matters. I still write code most days. I expect I always will. When I can, I'm outdoors, where the patience and the gratitude get topped back up.