Yuzzit is launching Reframe, an AI-assisted feature that converts live and recorded footage into vertical and other social-media layouts. Instead of applying a fixed crop, the system analyses frames, tracks moving subjects and adjusts the composition for the selected output format.
When two speakers need to remain visible, Reframe can create a stacked split-screen layout and shift emphasis toward the active speaker. The feature sits inside Yuzzit's existing clipping, subtitling and publishing workflow for destinations such as TikTok, Instagram, YouTube Shorts and X.
Why it matters
Vertical delivery often requires a second editorial pass immediately after a live event. Integrating reframing with live clipping can reduce turnaround and keep important action or speakers inside the mobile frame, especially when the original director did not compose for 9:16.
Automated subject tracking is not the same as editorial framing. Scoreboards, lower thirds, reactions and off-ball action can be important even when a model follows the main subject. Yuzzit has not yet published latency, supported input limits or accuracy testing for complex scenes.
