Making a 30-second video is disproportionately hard. You either hire a designer — $5,000, two weeks — or you open a template tool and become the designer yourself. Or you try GenAI and get a hallucinated video with your company name misspelled.
There's no option where you just say what you want and get something that looks like it was professionally done.
Workshop is that option. You write intent — a mood, a message, a layout — and a compiler handles the typography, timing, animation physics, and color theory. Same input, same output, every time.
The key insight is that a single mood setting encodes the hundreds of micro-decisions a motion designer makes per video. Change one word and the entire video transforms — easing curves, palette, spring physics — all recalculated from a single semantic input.
The same infrastructure serves a marketing manager writing a quarterly update and an AI agent generating a thousand personalized videos at scale. Because the input is just text, any LLM can write it. Because the compiler is deterministic, every output is brand-safe.
We believe the hardest part of making a video should be knowing what to say.