Why we built an orchestrator instead of a timeline
Every video editor starts with a blank timeline. We started with a blank prompt — and a planner that figures out the rest.
The first version of EditorBuddy had a timeline. Tracks, clips, a playhead — the shape every video editor has shared since the 1990s. It worked. It was also exactly as slow as every other editor, because a timeline asks you to do the thinking: where to cut, what to caption, which take was the good one.
We deleted it. Not the editing power underneath — the timeline as the entry point. What replaced it was a single text box and a planner we now call the orchestrator.
The job is a graph, not a list
When you type “turn this podcast into 8 shorts with captions,” the orchestrator doesn’t run a fixed script. It reads your prompt, looks at the source file, and plans a dependency graph: transcription has to finish before captions can start, but cleanup and scene analysis can run at the same time. Some of those nodes fan out — eight shorts means eight parallel reframe-and-caption jobs, not one job repeated eight times.
That graph is the actual product. The timeline, if you still want one, is a view onto the result — not the place where decisions get made.
A timeline asks “what do you want to do?” one click at a time. A prompt asks once, and the plan does the rest.
Why this matters more than it sounds like it should
Most “AI editing” features bolt a model onto an existing timeline: an auto-caption button here, an auto-cut button there. Each button is its own island — you still hold the whole project in your head, stitching their outputs together yourself.
An orchestrator means the islands talk to each other before you ever see them. Cleanup’s filler-word cuts feed the caption timing. Scene analysis tells the clip-finder which moments have faces in frame. None of that coordination is your job anymore.
What we kept
We didn’t throw away manual control — we moved it. You can open any node in the graph, see exactly what it decided and why, and override it. The difference is that “good enough, ship it” is now a real option, because the first pass is already a complete, coordinated cut — not a pile of raw footage waiting for you to start.
That’s the bet: most of the time, the plan is right. The rest of the time, you’re editing a draft instead of starting from zero.