Have you ever counted how many days it takes to get a single product spec approved? You ask a designer for wireframes, run another round of feedback, then go back to the engineers to ask what it will actually look like. One product manager at Stripe wrote his own tool that compresses all of that waiting into two minutes. Not Figma, not some third-party prototyping app — a clickable prototype built directly from the company's internal design system. The tool is called Protodash. For something built by a person who doesn't code for a living, it carries a surprising amount of weight.
Look closely at what actually happened at Stripe
Stripe's Owen Williams built the tool through what's come to be called "vibe coding": telling an AI model what you want, steering the code it produces, and revising as you go. The finished Protodash takes Stripe's internal design system as input and generates a clickable, interactive prototype in about two minutes. That work used to take a designer hours, sometimes days.
The speed isn't the point, though. What matters is that Protodash runs on Stripe's own design system. Where a generic mockup tool produces screens that merely look plausible, this one produces screens built from the same components as the real product. A prototype made by a PM ends up visually almost identical to what engineering would ship. The stakeholder question — "but what will it actually look like?" — simply disappears.
The story surfaced through Lenny's Newsletter in the first half of 2025. It was framed as a productivity hack, but look closer and it's more than that. Two facts sit side by side: a PM can now produce high-fidelity prototypes without a designer's help, and the tool is already in real use inside the company.
Here is the one thing that makes this different
Until now, a PM "making their own prototypes" mostly meant taking a Figma basics course or sketching simple flowcharts — and even that came with a steep learning curve. Even with low-fidelity tools like Balsamiq, the question "how will this actually be implemented?" never went away, and you couldn't answer it without a designer and an engineer signing off.
Protodash changes that structure itself. It proves, with a working product, that the job of making prototypes doesn't have to belong to designers. More important still: the person who built it isn't a professional developer but a PM. The person closest to the problem built the solution.
That raises a question. Now that AI has started absorbing skills and knowledge, the skill of building prototypes and the knowledge of a design system are no longer a PM's bottleneck — AI can handle a large share of both. So what's left for the PM? The ability to define a problem precisely, the judgment to know when and how to deploy a tool, and the will to fix an inconvenient reality with your own hands. That isn't skill, and it isn't knowledge. It's closer to attitude.
If you break competence into three axes — skill, knowledge, and attitude — AI tools are replacing the first two fastest. Attitude, by contrast — the refusal to look away from a problem, the willingness to try first even when the result won't be perfect — remains the territory AI can't copy. Owen Williams didn't build Protodash because he's a brilliant coder. He built it because the decision came first: "I'm going to fix this annoyance myself."
Of course, reading this as a shrinking of the designer's role is an oversimplification. What Protodash produces are prototypes for confirmation and alignment. User experience design, decisions about the visual system, accessibility review — all of that sits outside the tool's scope. If anything, designers are freed from low-fidelity review meetings and can spend their time on harder decisions. Roles aren't disappearing; each role is moving to a depth AI can't reach.
What this means if you're a solo founder or product planner
The first question is this: in your current workflow, which of the steps where you "hand things off to someone else" are actually necessary?
Many solo founders and independent product people outsource prototypes, proposal mockups, and landing-page drafts to contractors or collaborators. If the reason is "I don't have that skill," it's worth checking whether that premise still holds in 2025. Tools like v0.dev, Bolt, and Lovable already generate component-based UI without any design background, and Figma's AI features are rapidly evolving to fill in wireframes. A lower technical barrier doesn't mean you're now obligated to build everything yourself — it means you finally have the option to.
The second question runs along a different grain. The core of Protodash is that it uses Stripe's design system as input. Differentiated output appears only when you connect a company's own assets to an AI tool. If you run a one-person business or a small team, ask yourself: "What is my business's design system — the recurring language, tone, structure, and formats?" Define it, feed that context to your AI tools, and the output becomes far more consistent — whether it's a product spec, a proposal, or a content format.
The third thing to examine is how you use AI tools. If you're using AI today, are you using it as someone who receives output, or as someone who defines the problem and orchestrates the tool? Owen Williams built Protodash the second way. It wasn't a simple request like "make me a prototype in two minutes." He designed for himself what inputs were needed and what shape of output would solve the problem, then used AI as the instrument that executed that design. The difference doesn't stop at output quality — it determines whether you're the person controlling the tool or the person depending on it.
I won't list out specific things to try. Instead, one suggestion: within the next week, use an AI tool to produce one deliverable you would normally have handed off to someone else. It's fine if the quality is rough. That experience will sharpen your judgment about what to delegate next time and what to do yourself — far more than any list could.
What makes this tool — quietly used inside Stripe before the world heard of it — interesting isn't the speed or the technology. It's the attitude: the person who found the annoyance built the fix. Whatever your role, the most useful capability in the AI era still starts there. Skill can be borrowed and knowledge can be searched, but standing in front of a problem and saying "I'll take a crack at it" is something only you can do.
