decorative

Confluence: Notes on AI, design, & frontend

A confluence is where two rivers meet and become one. They don't take turns. They don't hand off. They flow together, and what comes out the other side is bigger than either was alone.

For most of the last decade, design and frontend didn't work like this. They worked like a relay. A designer made screens in Figma. A frontend engineer rebuilt them in code. The handoff was a meeting, a spec doc, a back-and-forth on spacing and edge cases that nobody enjoyed. Half the work was translation between two people who could see the same thing but spoke different languages.

The split was practical, not principled. It existed because no one person could hold both halves of the work in their head fast enough. Now they can.

Six months ago my team looked like every other design and frontend team.

It doesn't anymore.

Designers are starting to ship code. Engineers are starting to ship without designers. People who joined as one thing are quietly becoming another. They just picked up the other half of the job because AI made it possible.

This is happening everywhere, not just at Rango. Vercel runs an entire Design Engineering team and writes about it openly. Linear's website is built by design engineers. Airbnb hires for it. The Browser Company hires for it. The companies setting the bar for product quality on the web have already merged the role.

Hiring is the part this has made harder, not easier. The fewer roles you have, the more taste each person needs. The skill we're now looking for is judgment in front of an empty Claude Code window, not just craft in Figma.

The push for this came from a client partner we've worked with for the last 8 months. He kept asking why design and frontend were separate phases when the work didn't actually need to be. We didn't have a good answer. Recent progress in AI is what finally let us close the gap.

What's true inside the team is starting to be true for clients too.

AI native partners run differently. The team compresses. The timeline compresses. The quality only goes up. Founders who figure out which kind of partner they're hiring see ROI behind every dollar. The ones who don't will keep paying for handoffs that don't exist anymore.

I'm not sure where this ends.

Probably with niche teams shipping more, and a handful of partners that figured it out early eating the lunch of the ones that didn't.

Sagar Ludhiyani

CEO

CEO

Published On

Share on

Next