wheresharvey
The license is not the finish line.
Most AI programs measure adoption. Seats activated, logins counted, a dashboard that goes green the day everyone gets access. That's the easy number. It's also the wrong one.
Access is not capability. Capability is not readiness. You can hand someone a tool and watch them use it worse than they used the thing it replaced, and the dashboard will still say success.
I build the layer underneath the license: the curriculum, the judgment, the muscle memory that makes the tool actually change how work gets done. Technical AI capability times human readiness. Multiplicative, not additive. Zero on either side of that equation and the whole thing collapses, no matter how good your model is.
At MillerKnoll that's meant building a design system for how people learn AI (not just where they click), a heritage site that argues Florence Knoll and Charles Eames were already doing this work with different tools, and a sustainability model that tracks what a chatbot actually costs the planet per query (because "it's just software" is a story we tell ourselves, not a fact).
None of this is about the technology being new. It's about whether the people using it are ready to.
What would it take for your organization to actually be ready?
What I build
Featured work
- 01 AI Alphabet Twenty-six heritage entries that argue MillerKnoll design philosophy already contains the answer to AI literacy.
- 02 AI Toolkit Site An Astro-built enablement home with a token design system, UX audit rigor, and living curriculum.
- 03 AI Environmental Baseline GHG Protocol Scope 3 measurement of Copilot usage, a governance gap finding, and a bio-PU teaching case.