Emil Kowalski wrote two components you have used today without knowing it, and built a career on knowing when not to animate

There is a category of software fame that has nothing to do with being known. Emil Kowalski is in it. Sonner is the small stack of notifications that slides up from the corner of the screen when something saves. Vaul is the drawer that rises from the bottom of a phone. Between them they were downloaded more than ninety million times last week, which means his work is running inside a startling share of the React apps on earth, and almost nobody using those apps could tell you his name.

That is the design engineer's version of success, and he has it in a purer form than almost anyone. The components are opinionated, which is his word for them, and the opinions are correct: they look right, they move right, they get out of the way. People install them because the alternative is spending a week trying to reach the same place.

Vercel, then Linear

The résumé is short and it is the right two names. The design team at Vercel, where the bar for how an interface should feel is set unusually high, and now the web team at Linear, which is the company that made an entire industry of product people notice that software can be fast and beautiful at the same time. If you wanted to pick two places to learn taste by osmosis, those would be the two.

The through-line is motion. Not the kind that draws attention to itself, the kind you only notice when it is missing. A drawer that follows your finger with the right resistance. A toast that arrives and leaves without ever making you wait. That is the thing he is good at, and it turns out it is teachable.

Thirteen thousand people, ten days at a time

Animations on the Web is the course, and it is run like a product rather than a download: enrolment opens for ten days, closes, and reopens the next year. More than thirteen thousand designers and engineers have completed it. Four modules take you from easing curves and spring physics, through CSS and the Motion library, to the module that matters most, which is about the difference between an animation that works and one that is actually good.

The whole thing rests on one idea he keeps returning to: the most important decision in animation is when not to. Every other animation course teaches you to add motion. His teaches you where it belongs, which is a much smaller set of places than beginners think, and the restraint is the entire craft.

There is a second course now, AI for UI, in early access. Same instincts, pointed at the tools every design engineer is suddenly expected to use. If the first one is any guide, it will fill in ten days.

I care deeply about how they look, feel, and behave.

Emil, on animations, from the course he built around that care