EchoFrame

About this project

A browser-based video editor that treats AI operations as first-class editing tools. Instead of exporting to a separate service, a user can request captions or automatically extracted clips from inside the editing timeline. A React interface runs the editing experience, while a FastAPI backend coordinates speech-to-text and language-model calls that produce the effects. The engineering challenge is making slow, asynchronous AI work feel responsive within an editor, and mapping model output back onto precise points in the timeline.