A free 10-page field manual for the first ninety days of any serious AI program. Governance, decision routing, board reporting, and the deployment gates that get models past legal.
Four working tools for the first ninety days. Governance, decision routing, board reporting, and deployment gates.
Templates a CEO, CTO, and board can read in the same room. Use them as starters — mark them up, argue with the defaults.
Twelve items every mid-market AI program needs before the next pilot. SR 11-7 calibrated for non-banks — the framework that governs model risk at every U.S. bank, scaled down to what a 50–500-person company can actually maintain.
Where the model decides, where the human decides, and where you don't deploy at all. Most production AI failures aren't model failures — they're routing failures. This matrix forces a team to draw two axes and place every decision in a quadrant.
One page. Four sections: Value, Risk, Velocity, Capability. The numbers a CEO should be able to read to a board in under five minutes — not a deck, not a narrative, a scorecard.
A model is "deployed" when it has passed five gates, in order, with named owners signing each one. The unglamorous fix that unblocks more AI programs than any model improvement ever will.
The manual is a sanitized version of what I bring into Fractional CAIO engagements. Three credentials are doing the work in every page.
Direct-hire VP at a top-five U.S. bank, leading transformation across 30+ engineering teams under federal model risk governance.
Shipped this year alone — trading signal engines, biometric dashboards, executive intake tools, and more. Hands-on builder, not just an advisor.
Senior delivery lead at a major U.S. airline engagement: 120 teams, 1,200 practitioners, real-world AI rollout at the operational scale most consultants only describe.
I take on two new Fractional CAIO engagements per quarter. Mid-market CEOs, PE-backed portcos, Series B/C tech companies. Twenty hours per week, ninety-day initial term, value-based pricing.
One field note per month — a tactical observation from inside an active engagement, sanitized and useful. No newsletter spam, no automated drips.
No deck. No pitch. You'll know in the first ten minutes whether it's a fit. If it isn't, I'll tell you who I'd send the work to instead.
Book a 20-min call