We started with a crop
Every placement hands you a different box. Most tools render one film and squeeze it into all of them. We build to the box instead.
The problem we kept hitting
Our team ran paid acquisition for consumer and B2B products. The bottleneck was never the idea. It was taking one good cut and reformatting it for nine placements, then discovering the caption sat under the TikTok UI and the hero shot was cropped out of the Reel.
Generic video tools made it worse. They produced one polished master and left the reframing to whoever had After Effects open, which is where most of the week went.
What we built instead
Seedand treats the placement spec as the first constraint rather than an export setting. The hook is written to the duration cap, the subject is framed for the aspect ratio, and captions sit inside the safe zone before anything renders.
You give it a product photo, a script, an AI UGC actor, or footage you already own. It gives back concepts that are already the right shape.
How we work
We are a small, remote team. We publish our platform limits openly on the limits page so you can check our work, and we do not train shared models on customer briefs.
If that sounds like a place you would want to work, we are hiring.