← All speakers

Cova Lee

Conference affiliation: DevX and Tech Creator · OpenAI · 2026

Cova Lee is a video creator whose work combines visual design, music and custom editing tools. After building a full-time practice making short videos for Instagram, YouTube Shorts and TikTok, she joined OpenAI’s developer experience team. Her 2026 speaker profile credits her with an audience of 500,000 across platforms and collaborations with Adobe, DeepMind, Meta and Notion.

On this page

Cova’s early creative breakthrough came from her graduate-school dorm room. Using Photoshop’s generative fill, she transformed blank walls into a backdrop with lights, shelves, a window and a robot. She selected areas of the image, described what she wanted and chose among generated alternatives. She credits those experiments with helping her establish herself as a full-time creator: an ordinary room could become the setting for a much more elaborate visual idea.

Keeping the edit in her hands

In her workshop with Sarah Chieng, Cova explained why she often prefers an editable timeline to an automatically rendered video. Access to individual clips, titles, music and effects lets her keep adjusting the story. A rendered output can be convenient, but revising it may require another prompt and another render.

Her custom sound-effects pad makes that preference concrete. Its nine buttons insert sounds such as clicks, shutters and whooshes into an editing track. Cova can use it herself or have an assistant operating her computer use the same controls, then return to change the result. The tool reduces the work of placing sounds while preserving her ability to choose and revise them.

She applies the same approach to recurring editing tasks, maintaining instructions for operating Premiere and updating them as she finds inefficiencies. Those instructions cover steps such as sound effects, color grading and captions, allowing an assistant to identify unfinished work without requiring her to reconstruct the process each time.

References, experiments and a personal point of view

Cova uses existing work to make creative direction specific, and builds tools to test choices before committing to a shoot:

  • References that communicate style. Her visual swipe file collects videos, images and other inspiration. She also asked an assistant to study 50 of her Instagram videos and turn their recurring visual choices into a Figma design system. Her own work becomes a reference for subsequent production.
  • Experiments with titles and thumbnails. To learn YouTube packaging, she built a research tool for comparing videos, then added a simulated homepage to see how a proposed thumbnail looked among neighboring uploads. An outlier filter finds videos that substantially exceed their channel’s usual view count, helping her study unusual creative choices rather than audience size alone.
  • Music chosen before filming. Cova often picks music before making a video because she wants sound, images and text to work together. Music helps shape the piece from the beginning, rather than arriving only after the visual edit.

Her move into developer experience connected this creative practice with user education and product feedback. At OpenAI, she described creating content, talking with users, relaying feedback and helping address bugs. Her personal videos also formed part of that work, reaching people beginning to build with software beyond traditional programming roles.

Cova argues that a distinctive point of view needs experiences beyond the internet. Deliberately choosing what to do, noticing what draws her attention and developing interests outside technology supply material that references alone cannot provide. Her ambition to broaden who can become a developer follows that emphasis on personal agency: more people should be able to turn their own ideas into things they can make.

1 conference talk

Loading more talks and highlights…

References