Our team now runs 2X more product experiments with the new AI design process I created. Here is how.

2026 · SEOmonitor

Every project now gets a working prototype. Hypotheses are tested on the real thing.

Overview

Every project now gets a working prototype. Hypotheses are tested on the real thing.

We unblocked granular UX adjustments that help achieve business goals.

AI now provides an additional safety layer in identifying complex corner cases.

The same prototype is used as a handoff artifact for AI-driven development.

Also, I built a Chrome extension that mounts prototypes into the live product to get the real feeling of the project.

When the dev team adopted AI, we ended up with a lower delivery rate than before. I stepped in to solve it.

The system I created sits in its own repo that replicates our design system.

One command runs a pipeline of three skills, from context to shareable prototype.

Design principles were integrated as an additional knowledge layer.

This way, first-pass AI accuracy was enough to make the entire thing work.

That's how I made AI bring real value to the team.