Introduction
What VisoTalk is and how a document becomes an explainer.
VisoTalk turns an explainer — a product walkthrough, a code deep-dive, a book summary, a training module — into a structured, versioned document, the VisoTalk IR. The same document:
- plays in the browser as an interactive explainer: seek by sentence, branch, tweak parameters, click objects;
- exports to video: MP4, vertical shorts and subtitles, through a deterministic headless-Chromium renderer;
- is edited by agents and humans alike: agents write and patch the IR, and people use views that write back to it.
Why
Agents are already good at content: research, scripts, structure. What they are bad at is production — keeping narration and visuals in sync, laying things out without collisions, keeping one visual language, handling media licensing, and producing a first cut with no broken frames. VisoTalk is the production layer.