What Protogen Is

There's a pattern I keep seeing. A C-level product leader pulls me aside at a conference or on a call and says some version of the same thing: "I expected my design team to be further along by now."

They're not talking about tools. They're talking about a posture. Their engineers adopted Copilot, then Cursor, then agent-based workflows. Their data teams are building with AI daily. And their designers — the people who are supposed to be closest to the human experience — are still working the way they worked in 2023.

This isn't a failure of individual designers. It's a failure of how we've been framing the transition.

Most AI training for designers looks like this: here's a tool, here's what it does, here are some prompts to try. Go be more productive. It doesn't work. Not because the tools aren't good — they are. But because the change isn't about tools. It's about how you work. It's about your relationship to the material, your process, your instinct for when to reach for AI and when to trust your own judgment. You can't teach that in a workshop about prompt engineering. You have to teach it the way you teach any craft — by doing the work, in the new medium, with guidance from someone who's already been through the transition.

That's what Protogen is.

Protogen is a training program I created at Slalom to bring designers into the agentic era. Not by lecturing them about AI, but by putting them in front of the material and letting them build.

A typical session looks like this. A room full of designers — none of whom would call themselves developers. We open VS Code. We open Claude Code. And I say: build something. Anything. You have an hour.

The first fifteen minutes are always the same. Nervous laughter. People poking at the terminal like it might bite them. Someone asks "wait, do I just... type what I want?" Yes. That's exactly what you do.

Then something clicks. One person asks Claude to build a Pokédex — a searchable database of Pokémon with images and stats. They have a working app in twenty minutes and they're grinning like a kid. Another person builds a personal data management tool — something they've wanted at work for months but could never get prioritized in the backlog. Someone else makes a native mobile prototype for a side project they'd been sketching in Figma for a year. Real things. Things that work. Things they can show people.

The absorption in the room is total. People stop checking their phones. They stop talking to each other. They're leaning into their screens because they're finally touching the thing, not describing it. I've run this session dozens of times now, and that moment — when the room goes quiet because everyone is deep in the material — never gets old.

But here's what I care about more than the output. It's what happens to people's sense of what's possible. That's the first shift Protogen is designed to trigger — an identity shift. Most designers have spent their careers on one side of a divide: the people who envision things, and the people who build them. Protogen puts them on both sides at once. You're not a designer who's learning to code. You're a designer who can now build. That distinction matters. It changes what you think your job is. It changes what you think you're allowed to do.

The second shift is craft. Building with AI isn't like building by hand, and it's not like handing specs to a developer either. It's a new mode that requires new instincts. You have to learn when the agent is confidently wrong versus usefully wrong. You have to learn to read code well enough to know if the architecture makes sense, even if you couldn't write it yourself. You have to learn that the first output is almost never the right output — and that the interesting work is in the iteration. Prototyping at almost no cost. But working in code requires different instincts than working in design tools. You have to learn to think in systems, to debug with intent, to iterate on the real thing instead of a representation of the thing.

The third shift is collaboration. When designers can build, the relationship with engineering changes fundamentally. You're not handing off specs. You're not arguing about feasibility in a sprint planning meeting. You're sitting side by side, working in the same material, making decisions together in real time. I've seen designers at Slalom start creating feature branches, with lead devs running PRs to take that code into the main codebase. Six months ago, neither side would have believed that sentence.

There's a reason I built Protogen and didn't just write a blog post about it. The gap between design teams that adopt agentic workflows and those that don't is going to become unbridgeable. Not in five years — in the next twelve months. The teams that learn to work in this new way will move at a fundamentally different speed and produce fundamentally different quality. The teams that wait will find themselves unable to catch up, because the gap compounds.

This isn't about replacing any part of what designers do. The empathy, the systems thinking, the craft, the ability to hold a user's mental model while navigating organizational constraints — all of that is more important than ever. What Protogen does is give designers the fluency to apply those skills directly, in the material, at the speed the work now demands.

Every generation of designers has faced a medium transition. Print to web. Web to mobile. Static screens to living systems. Each one required letting go of old muscle memory and building new fluency. This transition is bigger than any of those. But the pattern is the same. The designers who lean in, who get their hands in the material, who tolerate the discomfort of being a beginner again — they're the ones who'll define what design looks like on the other side.

That's what Protogen is for. Getting designers to the other side.