I Run Data Center Operations for One of the Largest AI Fleets in the World. I Also Use AI Every Day.
Most executives at my level talk about AI strategy in meetings and then go back to spreadsheets and slide decks. I went the other direction — I actually built AI into how I work.
I don’t prepare for meetings the old way anymore
When I prep for an operational review, I have an AI agent pull data across the multi-state region, analyze trends, flag anomalies, and draft the narrative. I still make every call. But I’m making those calls with better information, faster.
When I walk into a legislative briefing at a state capitol, I’ve already had an AI agent research every legislator I’ll meet — their voting history, committee assignments, public statements, and where they stand on the issues that affect my sites. What used to take a policy team days to compile, I have in an hour.
I model scenarios in minutes, not days
Workforce planning at gigawatt scale is genuinely hard. How many engineers do you need for a site that will grow from 500 MW to 2 GW over five years? What’s the crossover point where insourcing beats contingent labor? What happens to your staffing model if diagnostic accuracy improves by 10%?
I used to wait for analysts to run these models. Now I describe the scenario to an AI agent, feed it the constraints, and iterate in real time. The quality of the output depends entirely on the quality of my thinking — the AI just removes the friction between having an idea and testing it.
I use AI to improve how my teams use AI
I sponsor AI initiatives across our operations organization — automated repair systems, troubleshooting agents that work alongside technicians, LLM-powered quality scoring of repair tickets. But I don’t just govern these programs from a distance. Because I use AI tools daily, I have real intuition for what works, what doesn’t, and where the failure modes are.
When someone proposes an agentic AI system for automated server repair, I can pressure-test it because I understand how these tools actually behave — the hallucinations, the edge cases, the places where human judgment is still irreplaceable.
What I’ve learned
Most leaders are delegating AI to their teams instead of using it themselves. That’s backwards. The person with the most context should be the one interacting with the AI.
AI doesn’t replace judgment — it compresses the time between question and answer. The hard part was never the analysis. It was getting the right information assembled fast enough to make timely decisions.
Fluency compounds. Every week I use these tools, I get better at knowing what to ask, how to structure problems, and where to trust vs. verify. Leaders who wait to adopt are falling behind in ways they can’t see yet.
The next generation of operations leaders will be AI-native. Not AI-curious. Not AI-aware. Native.
The gap between leaders who adopt AI tools and leaders who delegate them is going to be the defining career differentiator of the next five years.