MKFrame
Editing Automation

Meet the AI Edit.

This is AI-automated editing at work, on a real screen.

AI handles repetitive editing,
so people can focus on judgment

MKFrame designs and runs workflows that automate content editing with Claude and Codex. From rough cuts to captions, color grading, generative graphics and short-form, each stage is handed to AI while people focus on decisions and review.

A simplified illustration of automated edits stacking up on the timeline.

How we work

People decide, AI processes, people review.

  1. Direction and decisions

    The producer sets the editing direction and standards.

  2. AI processing

    Claude and Codex handle the work stage by stage.

  3. Review and approval

    The producer checks and approves the result. Corrections become the standard for the next job.

Research directions

Four directions we are researching to widen what can be automated.

Style quantification
Turns an editing style into numbers so AI can follow it as a standard.
Editing-stage automation
Moves editing stages, from rough cut to short-form, into AI work one by one.
Generative graphics
Combines image generation with code-driven motion to produce video graphics.
Automated review and feedback learning
Automates checking of the output and feeds corrections into the next job.

Curious how the automation works?

Feel free to ask about the workflow we use on 명승은TV.

Contact us