Notes from the floor.

Engineering posts, creator stories, product updates from the team.

How our 8-stage AI pipeline works

Ten DB states, eight working stages, five UI tiles — the actual handoffs between them, with the vendors that do the work.

We don't ship a retention model. The hook stage gives you swappable openers — and why that beats predictive scoring.

Two tables, one approval gate, and the single field that decides whether your scenes look like the same video.

Three TTS providers wired — ElevenLabs, Inworld, MiniMax. No custom-trained models. When to pick which.

Credits, render time, queue depth — what it costs to scale, with no fictional creator success story.

What our shot planner actually does — semantic beats, style guides, reference images — and where prompts still leak through.