— Insights
Essays on the parts of AI transformation that don't get written about — the structural conditions that determine whether the work lands, the artifacts that hold engagements together, and the engagement structures that align incentives. Written for executives who have already heard the strategy decks.
An AI strategy names ambitions. AI Decision Architecture names the decisions — who decides what, on what evidence, with what gates. Why the second is the thing boards are actually missing.
Most AI backlogs are lists of ideas ranked by enthusiasm. A portfolio map scores them on value, feasibility, and risk — and turns 'which idea is exciting' into 'which initiative gets funded next.'
A demo proves a model can produce an output once. Diligence proves there is a defensible business underneath. The red-flag patterns that separate genuine AI depth from a prompt and a logo.
The first 90 days decide whether an AI initiative compounds or quietly dies. A phased, gated plan — Diagnose, Design, Build, Embed — with owners, gates, and the off-ramps most plans never write down.
Most AI pilots don't fail because the technology was wrong. They fail because the conditions for success were never put in place. Five recurring patterns, and the structural fix for each.
The one-page artifact at the heart of every Diagnose phase. What's in it, why it works, what it looks like — and the situations where it's the wrong tool entirely.
A piece for CFOs and CSOs evaluating whether a hybrid advisory/venture engagement structure makes sense. What the mechanism is, when it applies, and the situations where the standard advisory structure is the right answer.
Almost every enterprise now has AI initiatives. Most are AI-enabled — features bolted onto the old model — when they were funded as if they were structural. The eight questions that tell you which ones actually compound.