Find the moment. Make the clip.
Connecting transcript-based selection to clip rendering in a focused local tool.
The problem.
The useful moment in a long recording is often easier to find in its transcript. Turning that selection into a clip requires a connection between the text, source timing, and rendering process.
What I built.
I built EventCut to connect recording intake, transcript-based curation, and clip rendering. A selection maps back to the recording, and a processing worker produces the selected video. The current version runs locally for a single user.
The decisions that matter.
Start with the transcript
Use the spoken material as a way to identify a passage worth clipping.
Carry the decision into the render
Keep the source selection and timing connected to the processing workflow.
Focus the current product
Center the local version on making clips, with a clear path from source to output.
The result.
Select a passage in the transcript and the tool renders that clip from the original recording, timing intact.
Built with.
- Next.js
- React
- TypeScript
- FFmpeg