Before I built AI systems, I spent three decades architecting brands and packaging for some of the world’s largest companies across five continents.
The problems I solved then and the problems I solve now are more similar than most people realize.
When you manage a global brand portfolio, you quickly learn that creativity is not the scarce resource. Consistency is.
The same is true with AI. Anyone can generate impressive one-off outputs. The real challenge is creating systems that remain reliable, on-brand, and safe across thousands of interactions.
In packaging and brand architecture, we didn’t rely on “brand guidelines” alone. We created strict hierarchies of claims, visual systems, and regulatory constraints that were non-negotiable.
This is exactly what I now encode into AI systems as governance contracts — hard rules the agent cannot break.
I see companies making the same fundamental errors with AI that they once made with global brand rollouts: chasing speed and local adaptation while slowly destroying the coherence and trust that made the brand valuable.
The technology has changed. The underlying discipline has not.