Arte Mentis

Independent AI consulting · San Diego

Know what AI is actually worth to your business before you spend on it.

I help owner-led businesses find where AI genuinely pays off, hand the tedious work to tools that can actually carry it, and check vendor claims before a contract is signed. I published a benchmark most models failed. The skepticism is part of the service.

Pinus torreyana. The rarest pine in North America. Scarce things, like your good hours, are worth protecting.

Recent work

≈45× recall

Built an address enrichment service for a B2B sales team that raised tenant recall roughly 45 times on the test batch (from 8 businesses found to about 368). It now runs nightly in production with no manual involvement.

7 models tested

Benchmarked seven language models against fixed thresholds for data extraction, mathematical reasoning, and executive summarization. One passed.The framework and results were published in Towards AI.

200+ trained

Onboarded and trained 200+ distributed professionals for an AI evaluation firm, cutting performance variance across the group by 40 percent.

Services

Assessment

Focused AI Assessment

A short, structured look at your operations that ranks where AI would actually pay off, what it would cost, and what to skip. You get a written report with numbers you can act on, whether or not I do the work that follows.

Adoption

AI Adoption Sprint

Your team already has the tools. This gets them used. Hands-on training built from a decade of teaching people to work with complicated systems, with materials your team keeps and a plan that does not depend on me staying.

Diligence

Vendor Second Opinion

Before you sign an AI contract, an independent read on what the product actually does, what the deck is really claiming, and whether the price matches the value. Fixed fee, delivered in writing.

Also available for fixed-scope builds and ongoing advisory. Every engagement is scoped in writing before it starts. See the full rate card.

About

I’m Matt Simmons. I test AI claims before believing them. In December 2025 I built a benchmark that put seven language models through JSON extraction, mathematical reasoning, and executive summarization against fixed thresholds. One passed, and Towards AI published the results.

Behind that sits a decade of teaching people to use complicated tools. Seven years as a graduate instructor at UC Santa Cruz, twenty-plus courses, more than a thousand students across every level of technical comfort. Earlier, I managed twenty-plus institutional accounts for a B2B education platform, including Harvard and MIT, with a 95 percent renewal rate. I hold a Ph.D. from UC Santa Cruz, which in practice means I read what a vendor is actually selling rather than what the deck appears to say.

My stance on AI is simple: it shouldn’t do everything. It should do the tedious things (securely, on your terms) so you get back time for the work only you can do. More on where I stand.

I’m based in San Diego and deliberately work with a small number of clients at a time.

Contact

The first conversation is a fit call, not a pitch.

Tell me what you’re trying to figure out, and I’ll tell you honestly whether I’m the right person for it.