Technology ID
TT-0042
Description
The essence of video editing is telling a story smoothly and fluently. Today’s video foundation models mainly learn to understand video, but are not trained to edit – e.g., selecting the right shots from massive footage to build coherent narrative flow. Zonic fills this gap with a learnable Agentic AI that reviews footage, proposes multiple story-driven rough cuts, and iterates with professional editors. By using editors’ final cuts and feedback as “reviews,” Zonic improves every project, reducing heavy human labor now and accumulating editing intelligence that will later benefit layman creators.
Content
Storytelling-First AI (Editing Decisions, not just Understanding): Selects and sequences footage to achieve fluent narrative flow
Editor-In-The-Loop Learning: Improves continuously from professional editors’ final cuts and feedback; “evolves every project”
Learnable Agentic AI (All-In-One, not a Toolbox): Instead of giving many separate AI tools and training humans to use them, Zonic integrates the workflow end-to-end and automatically learns skills from humans through interaction and outcomes — becoming more capable over time
Inventors
Applicable Industries
Potential Application
Technology Focus
TRL
Images
