I’m building Produca around a fairly ordinary starting point: someone has an idea for a video, but doesn’t want to spend the afternoon translating it into the vocabulary of several model interfaces.
They describe an ad, story, reel or animated poster, choose a direction, and add references where needed. The work between that brief and a video with sound is the part I’m trying to make easier.
Keeping the useful decisions
One good clip is a start. A second clip that forgets the product, character or setting creates more work. Produca groups those references into projects so they can be reused instead of described from scratch.
I’m interested in how much control to expose and when. A first draft should be easy to request. A revision should be specific: change this shot, keep that character, shorten the narration. If every correction becomes another large prompt, the interface hasn’t saved much effort.
Cost makes those decisions more important. An expensive generation is a bad place to discover that the brief was misunderstood. Scripts, references and intermediate previews give the person something to approve earlier.
This is a solo product, so I also have to think about operating it: provider failures, incomplete jobs, usage and support. I’ll share more concrete implementation notes as the work develops. This piece is about the direction, not a claim that all those problems are solved.