In 1997 a computer beat the world chess champion. What came next was more interesting: a human paired with a machine beat both. Mentaur builds the systems that work that way — and ships them.
The wedge
I sell this because I run this.
The AI market is full of things that have never carried real work. Polished demos. A waitlist. Screenshots of a product nobody has run on a Tuesday afternoon.
I run a multi-agent pipeline that has published novels and podcasts. I built and shipped a production AI receptionist that books real jobs on its own. I took roughly six hundred engineers through a tooling migration in a certified environment and have the adoption numbers to show what changed.
That's the difference. I'm not selling a thing I read about. I'm selling a thing I run.
Proof of work
Two systems I built. One room I changed.
The first two say I can do this. The third says it works on other people.
Writers' Room
A production multi-agent system: 15 specialized agent types, structured handoffs, quality gates. It has shipped published novels, a completed musical, and two ongoing podcast series. Not a prototype — a pipeline that produces finished work.
CallCatcherPro
A production AI receptionist built on the Claude API with Python, Vue.js, and AWS. Handles live inbound calls, natural-language booking, and inquiry routing for small businesses. Concept to code to answering the phone.
It works on other people
Everyone has sat through the inspiring session that changed nothing. Leading a ~600-engineer migration, I built a generative AI workflow that cut documentation-to-script review from days to hours — and the training moved the whole org: engineers actively trained up ~11×, pipeline runs ~39×, merge requests ~44×.