Every person on your team has a different "user" in their head — and nobody ever says it out loud.
A proto-persona fixes that without needing a research budget or weeks of interviews. It's not research. It's your team's assumptions, written down and agreed on, so everyone designs for the same person instead of five different ones.
Here's how I build one: ask AI to draft it first, strip out anything that sounds like a stereotype, get the team in one room to argue with the draft, then label it clearly as a guess — not a fact.
A proto-persona won't replace research. It'll tell you what to research first.
Does your team actually agree on who you're designing for — or has nobody checked?
AI can generate a dashboard. But can it give a product its own identity?
The first design was created by my client using AI a familiar dashboard style you've probably seen across countless AI-generated products.
I redesigned it to feel less generic and more intentional, with a distinctive visual identity, clearer hierarchy, and an experience built around meaningful celebrations.
The goal wasn't just to make it look better. It was to make it feel like its own product.
What do you think is missing from most AI-generated interfaces?