I didn’t set out to build a product.

Before COVID, the company I work for had more than twenty people on payroll. We lost them slowly, and not to layoffs — to arithmetic. Good people asked for raises we couldn’t give and benefits we couldn’t carry. They were right to ask, and they were right to leave when the answer was no. There was never a round of cuts. We just stopped being able to refill the chairs.

By 2025, four of us were doing the work of twenty. Not as a figure of speech — the same customers, the same books, the same infrastructure, the same deadlines, spread across four people and their evenings. Anyone who has run a shrinking company knows the shape of it: every month you get slightly worse at everything, and the people still there pay the difference in hours.

So I staffed the empty chairs the only way left. Not to replace anyone — there was no one left to replace. The roles had been vacant for years; the work had never stopped arriving. I built a roster of AI staff to hold them — the books, the marketing, the code, the infrastructure — under rules I could enforce rather than hope for: everyone drafting first, trust granted a rung at a time on work I’d watched, a gate no one on the roster can pass, and a ledger that records every permission I widen, with the reason.

It worked well enough that the four of us still have jobs. That sentence is the entire origin of this company.

Then it occurred to me that we are not rare. There are a lot of companies like ours — squeezed between what good people rightly cost and what the business can currently carry, run by a handful doing the work of many, told by every vendor to scale faster. This is for them. Not to grow headlong. To hold: stay open, keep the humans you have, and live to fight another day — including the day the arithmetic lets you hire again. A dead company employs nobody.

It replaces execution, not judgement — not relationships, not anything with hands — and the architecture is built so it can’t pretend otherwise. Everything irreversible, everything that risks the company, every dollar past the tier you set stops at you, at every trust level, enforced at the API.

Axioma Facere runs on its own roster because it was born on one. The failures reach us first, and the fixes ship because we needed them.

If your company is in the place ours was: this was built there.

— Joshua Holland