Bringing AI prototyping to the whole UX team

The short version

Our team prototyped in Figma, and Figma prototypes have limits. They can't easily show real data, logic, or the edge cases engineers need to see.

Watching Figma's Config conference, I saw where prototyping was heading and realized we already had access to Kiro, an AI tool that could get us there without buying anything new. I tested the idea with my manager, worked with our software engineering director to set up a shared code repository, taught the UX team to build working front-end prototypes with it, and built a team dashboard where everyone publishes their work. Today every designer on the team prototypes this way on every project, and the prototypes are part of developer handoff.

The problem

• Figma prototypes hit a ceiling. Real data, filters, logic, and edge cases are hard or impossible to show, so engineers had to interpret the gaps.
• Feedback on behavior came late. Users and stakeholders could click through flows, but not use something that behaved like the real product.

What I did

• Found the idea. Figma's Config conference showed where prototyping was heading. I realized we already had access to Kiro, so we could get there without new software.
• Tested it with my manager first. She had already experimented with it, and together we confirmed it was worth rolling out to the team.
• Set up the foundation with engineering. I partnered with our software engineering director to create a shared code repository designers could safely work in, and got the engineers on my own product teams on board with prototypes in handoff.
• Taught the team. I showed every designer on the UX team, including my manager, how to set it up and use it, and stayed the go-to person for questions.
• Made the work visible. I built a team dashboard where every designer publishes their prototypes, so product owners and engineers can find the latest version of any project in one place.

Key decisions

• Use what we already had. Kiro was already available to us, so there was nothing new to buy or get approved.
• Prove it before rolling it out. Testing it with my manager first meant the team adopted something that had already worked, not an experiment.
• Start where I had influence. I got the engineers on my own product teams on board first, then shared the setup with the rest of the UX team.
• Visible by default. The dashboard turns prototypes from files on someone's laptop into something the whole product team can review.

Outcome

• Every designer on the UX team now prototypes with AI, on every project.
• Prototypes are part of developer handoff, so engineers see real behavior, not just screens.
• One place for every prototype. The dashboard gives product owners and engineers a single place to review work in progress.

What I learned

• Leading isn't only managing people. One of the biggest impacts I've had on the UX team came from changing how we work, not from a title.
• Look outside, then use what you have. The idea came from Config; the win came from realizing we didn't need anything new to act on it.
• Make new habits easy to keep. Training gets people started; the dashboard is what keeps them going.