Long-form in. Editorial decisions out.
Turn long conversations into clips worth watching.
ContentAgent finds the meaningful talk inside podcasts, interviews, and classes, then turns it into a coherent cut with speaker-aware captions.
UP TO 03:00:00 SOURCE / 01M-20M OUTPUT / 9:16 + 16:9
Founder Sessions / Episode 18
SOURCE 02:14:32 / ANALYSIS READY
The decision that changed the business
00:42:18
Why consistency beats intensity
01:16:04
A practical way to protect focus
01:49:36
01 / Selection
Find the parts worth keeping.
The planner filters introductions, sponsor reads, filler, and dead space. It keeps complete thoughts and shows where every chosen range came from.
The Creative Practice / Full conversation
SOURCE 02:14:32
The decision that changed the business
Natural sentence and turn boundary
Why consistency beats intensity
Natural sentence and turn boundary
A practical way to protect focus
Natural sentence and turn boundary
The work gets better when the decision gets clearer.
02 / Structure
One coherent section, or only the moments that belong together.
Auto Best compares a continuous conversation against a natural compilation. It chooses one output plan, protects context, and avoids forced cuts.
Editorial rule
A complete thought is more valuable than hitting an arbitrary cut count.
03 / Speaker-aware captions
Every voice keeps its place.
Pyannote separates speakers, overlap uses distinct lanes, and the correction editor gives you a precise manual safety net.
What changed when you stopped chasing every opportunity?
I finally had enough attention to make one thing excellent.
That sounds obvious, but it is a difficult choice.
04 / Output
Choose the destination before the cut.
One source can become a one-minute vertical clip, a longer horizontal excerpt, or anything between.
Duration
01 / 02 / 05 / 10 / 20
Minutes
Frame
9:16 / 16:9
Shorts and YouTube
Built for long sources
Chunked transcription keeps three-hour videos practical on local CPU and disk. Processing status stays visible from upload through analysis.