Video · Automation
AI video editor
From raw footage to a considered edit.
Make the cut.EDIT → REVIEW → PUBLISH
THE STARTING POINT
Why I
made it.
I love filming. Editing is the part I keep wishing I could skip. After manually editing more than 100 videos, I wanted to build a system that could handle the repetitive work without flattening the choices that make an edit feel human.
What I built
An agent-operated local workflow for transcription, captions, colour and audio adjustments, silence cutting, rendering and review. A related publishing workflow prepares platform-specific copy and schedules content through Buffer.
My role
I defined the editing choices I wanted to automate, connected the tools and reviewed the output. When a filter looked wrong, I supplied older videos I liked as references and directed the agent to revise it for natural skin, richer colour and better blacks.
What works today
I have used the evolving editing workflow on around 20 personal videos. That follows 100+ videos I filmed, manually edited and uploaded myself. The automation is still being improved, with my review of the final look.
What it taught me
Good editing is not only cutting files. It is judgment: what should stay, what should disappear and what the viewer needs next. The system handles the mechanical work while I keep control of the idea, pacing, style and final result.
HOW AN EDIT MOVES THROUGH THE SYSTEM
- 01
Ingest
Bring in the source footage.
- 02
Understand
Transcribe and analyse scenes.
- 03
Plan
Review the transcript and choose the edit.
- 04
Render
Assemble clips and captions.
- 05
Review
Check the output and adjust.
The editor and description components worked on real batches of my content. They were designed to feed a broader ten-agent workflow that prepares Buffer drafts, while I keep the final approval.
A CLOSER LOOK
GOT SOMETHING IN MIND?
Let’s make
something happen.
A project, an experiment, a “could we build this?”
I’m happy to hear it.