Twenty-plus years across product, engineering, and operations — at companies you've used and startups you haven't heard of. These days I'm obsessed with one question: how do you work alongside AI agents without losing the judgment that makes the work yours? This site is where I think out loud about it.
Most expertise is invisible — it lives in the gap between the question and the answer. I'd rather make that gap visible. The studio on my homepage isn't a demo; it's my actual workflow, running live, including the messy parts.
AI agents can draft, cut, schedule, and summarize. They cannot decide what's true, what matters, or what I'd be proud of in a year. I automate aggressively and approve everything. The 5% only I can do is the whole point.
No gatekeeping, no course, no "DM me to learn the secret." Just an essay a week breaking down one more piece of how I work — while I'm still figuring it out.
This isn't a CMS. Essays from my current pipeline are drafted by an AI agent, scored against a rubric, and approved by me before shipping. The orchestration happens in code—migrations, concurrency guards, cost tracking—all live. If you're curious how the system works (the architecture, the tradeoffs, what's built vs. aspirational), there's a full technical breakdown.
Read the Architecture →I take on a small number of consulting engagements each year. If your team is ready to integrate AI agents into production workflows, let's talk.
Work With Me →Get essays on AI-assisted work