Problem
Decisions validated too late to be cheap
Design decisions were getting validated too late — after implementation had already started — because there wasn’t a fast, shared way to test an idea before committing real engineering time to it.
This wasn’t a problem specific to one team. It affected every designer across three different products.
Process
Pick the tool, then change the practice
I evaluated more than twenty AI-powered prototyping and design-to-code tools before selecting Dessn, then drove its adoption across design, PM and engineering.
The tool was the easy half. I led the rollout of prototyping as a practice across the entire design organization — not just my own team, but working with designers on products I don’t build for at all.
The goal was never to get people to use a tool. It was to move the moment of validation earlier, before it costs engineering time.
Technical depth
What the work actually involved
- Ran a structured evaluation across 20+ candidate tools before committing to one.
- Drove cross-functional adoption spanning design, PM and engineering.
- Established prototyping as a standard step before implementation across three separate products.
Outcome
It’s no longer something I have to actively champion. It’s just how the design org works now, day to day, across every product it touches — which is the difference between a pilot that fades and a practice that actually sticks.