Thinking in public on agentic AI and what it means for technically inclined leaders.
AI can help bury the gap between intent and implementation, but only if organizations let it reorder the status markers around software. Otherwise they'll replay the same transformation failures in a faster loop.
Builder-leaders spent careers choosing between building and leading. AI agents are ending that tradeoff — and it turns out the judgment they developed as leaders is exactly what working with agents requires.
The real risk in agent workflows is not that builders lose their skills. It’s that summaries and convenience make judgment feel optional.
Agentic AI is bringing governance back into the room. This is a personal reflection on why I’ve come to see the good version of it as an enabler, and the bad version as fear with a badge.
Why the employee metaphor produces better management instincts than the prompt metaphor when you're building agentic systems.
What it took to make three OpenClaw agents feel like part of the same team instead of three disconnected toys.
The RevOps agent on the team, built to own the quiet systems work that keeps a revenue organization from drifting into dysfunction.
The second agent on the team, built to turn scattered customer signal into usable account awareness for a small team.
The first agent on the team, built to own sales development work and test whether an agent could hold a real lane inside a revenue team.
What happens when join the team? I hired three agentic employees into my team at Replicated. Learn what they did and how it changed the way I think about leadership, leverage, and org design.
What if the thing that made you good before leadership is still the thing? When I stepped into sales leadership, I realized the inherited template was built for someone else. AI is making it possible for leaders to build again instead of choosing between abstraction and the tools.