Auto-Create Audiogram Social Clips from Each New Podcast Episode
An n8n workflow that picks the most shareable 30-second moments from a new podcast episode and renders them as captioned audiogram videos for social — turning one episode into a week of promo clips without a video editor.
- 1
Detect the episode
An
RSS Feed Triggerfires on a new episode and anHTTP Requesttranscribes the audio. - 2
Pick 3 clips with Claude
An
HTTP Requestto Claude returns strict JSON of three segments — start, end, caption — chosen for shareability. - 3
Render each audiogram
An
Item Listssplit then anHTTP Requestto an audiogram API renders each segment with waveform and captions into an MP4. - 4
Save for review
A
Google Drivenode stores each clip and aGoogle Sheetsnode logs the caption, so posting is copy, review, share.
Frequently asked questions
How are the best moments chosen?
The transcript goes to Claude, which returns the three most quotable, self-contained 30-second segments with their timestamps. Those are the clips most likely to stop a scroll — a strong quote, a surprising stat, a funny exchange.
How is the audiogram video made?
An audiogram API (or a headless renderer) takes the audio segment, a waveform style and the auto-generated captions, and outputs an MP4. The workflow passes each chosen segment and caption to it, then saves the clips for review.