Build log — December 2025

The build log is a monthly record of building AquaIntel: what shipped, what broke, what a customer said, and one number. Customers are never named — these come out of private calls and the people in them did not agree to be written about.

The month I started again. At the end of November I had taken out almost everything I built and asked whether one thing was left that a customer would pay for and that I could deliver myself — that story is here, and why I had stopped in the first place is here.

Shipped. One utility profile, entirely by hand. Gettysburg, Pennsylvania, a town of about ten thousand people. Board minutes, capital improvement plans and financial reports read end to end, turned into five opportunities I believed they would want solved in the coming years. It took about a week. The newsletter it was meant to feed never launched, because the profile turned out to be the product.

I also finally read what GWI and WEF had launched back in September — three months after deciding it had ended my company.

How I described the company that month, in my own words on a call:

At the moment, I'm not offering the software platform. I am offering this profile or intelligence report on a particular utility that I'm able to run on my back end.

Worth preserving. It is the smallest this company has ever been described.

What a customer said. Two conversations this month arrived by routes I could not have engineered.

The first was a former client who found the old website — which was still up only because when I decided to stop I forgot to take it down. He did not know the company was mine. He read what it claimed, went looking for whoever was behind it, worked out it was me, and got in touch. That call is most of the reason there was a December at all.

The second, two days before Christmas, was a utility veteran of thirty-two years who had just moved into equipment sales. He had written up a platform idea of his own and asked an AI whether anyone was already building it. It pointed him at me. He booked a call to tell me I had built the thing he had been sketching.

What he wanted was not the database:

The magic to make the sale is still in the personal contact... but it's having the insight and the ammunition of the background when you walk into the conversation. They're like, how did they know that?

Broke. The newsletter — in the way a plan breaks, rather than a system. It did not fail; it never launched, and the reason is a few paragraphs up. What is worth recording is that I did not act on it for another month. The swap had already happened by the end of that week of reading: the research had become the product, and the product had become unnecessary. I went on describing the newsletter as the plan anyway, to myself and to other people, long after it had stopped being one.

Systems break loudly. Plans break by sitting unchanged in your head while the facts move underneath them.

One number. One profile. One recorded conversation.