Before I built software, I ran the businesses.

I started on a restaurant floor at Jimmy Buffett’s Margaritaville inside Mohegan Sun, and stayed five and a half years — supervisor first, then operations manager and head of the beverage department. That meant the schedule, the staffing, the inventory, the pour cost, and a profit-and-loss statement I had to explain every month. A casino restaurant turns more covers in a night than most rooms do in a week, and none of it was run on software built for it. It was run on spreadsheets and on me.

From there, banking. I ran consumer lending operations at United Bank in Hartford — the day-to-day, the policies, the hiring and training — then cash management at Dime Bank, where the work was financial reporting and closing the books clean enough to survive an audit. Different building, same shape of problem: the system covered most of it, and a person covered the rest by hand.

Then a clinical practice, where I ran patient scheduling and treatment records under federal privacy rules. If you run a dental office or a family practice, you already know that work. After that a licensed insurance practice — reading policy documents to people who needed them explained, and working out what someone actually needed rather than what was easiest to sell.

And now supply chain at the appliance plant in town. Materials planning, bills of material, supplier performance, and the ordering logic that decides whether a production line runs on Monday. If a part is missing, the line stops — so the whole discipline is knowing what you are about to run out of, long before you run out of it.

Five trades. One problem, every single time: the software almost fit, and somebody was quietly holding the rest together.

That gap used to be permanent, because the software that would have closed it was priced for companies a hundred times the size. That stopped being true. I build those tools now, and the business that gets one owns it outright — codebase, data, accounts in its own name.

The AI half isn’t new either. It ran alongside the operations work the whole way, and it’s now formally part of the work — shaping the path, scope and implementations of AI across the company’s North America region. But it’s the leverage, not the point. The point is that a shop running on a spreadsheet can now have a real tool built around exactly how it works — no enterprise budget, no licence, no asking permission.

I wasn’t born here. I chose it.

See where that’s pointed — my focus on local small business →