I’m a strategist and operator based in Boston. I connect strategy to execution: I find where value is hidden in a business and build the systems, often AI agents, that release it. My background covers business development, founding a startup, venture, and international go-to-market work.

I start with the people doing the work, not the technology. I map how information actually moves, find where it stalls, and set how much the agent can do on its own based on each team’s tolerance for error. Then I pilot one real use case before scaling.

Yes. I build the first version myself with tools teams already use, like Copilot Studio, Power Automate and SharePoint. When something needs system access, security review or scale, I hand IT a full spec rather than an idea.

It depends on the department’s risk tolerance. Sales can absorb a draft that gets corrected; operations can’t. That line decides whether the agent acts, only suggests, or waits for a person to sign off.

The reasoning and cost of the problem, a data map of what the agent should and shouldn’t read, options to keep costs down, defined outputs, tested agent instructions, and a phased scope that can be approved one piece at a time.

Most of my work has been in shipping and logistics, education, startups and venture, and event tech. The approach carries over to any business where work runs on email, spreadsheets and what’s in people’s heads.

My blog about AI from inside a traditional industry, and about the difference between judgment and answers. I also write honest career advice for people early in their careers.