Field notes for the decision layer.

It took us the last quarter of 2022 just to find and build the team. From 2023 through 2026, that team built an AI practice inside ANVL Entertainment, our working film and television company. Self-funded. No vendor to please, no product to sell, no demo to keep alive for investors. Just the question every operator eventually faces: what is this actually good for, here, now?

We built a lot, and we built early. Script coverage. A development pipeline. A story diagnostic. Storyboarding — including video storyboards. Research tools that could make us subject-matter experts on any world a script touched. None of it was a wrapper on someone else's product. All of it was decades of producing and operating judgment encoded into instruments we actually ran — and that blend, expertise into tool, was the entire point. It cost us: real money, real time, every lesson paid for.

Which made the temptation to productize real. There was a window when packaging any one of these and taking it to market looked like the obvious recoup, and we watched others take that road. We passed. The read: the field was compounding monthly, so anything we shipped would be obsolete before the market could find it — and chasing a product would have traded the advantage we were building for a claim we'd have to keep defending. No defensible reason to ship. So we didn't.

We also killed pilots that worked. That's the part that surprises people. A pilot that demos well and a pilot that survives real use are different animals, and some of ours were the first kind wearing the second kind's clothes — tools that needed more supervision than they saved, builds whose edge would evaporate with the next model release. Working isn't the same as durable. We stopped them early, before they could quietly absorb budget and belief.

We stayed agnostic the whole way — fluent in every commercial LLM and in the open-source smaller models, committed to none, because allegiance is a cost you pay with your options.

We hacked our way through the moat ourselves — thrashed about, took on water, learned which crossings hold. That is the judgment DeepMoat brings to the decision on your desk.

— Keri Nakamoto, DeepMoat