what's in the export
- episode summary — three to five sentences pulled from the full transcript, written in your show's voice (configured once per show). edit in place before export.
- chapter timestamps with linkable anchors:
[00:04:12] the framing problem. these match the chapter markers podcast apps read, so apple podcasts, overcast, and pocket casts pick them up automatically. - cleaned-up paragraphs — the AI-tighten pass collapses repeated phrases, removes "you know" / "like" / "i mean" filler, and breaks runs of speech into paragraphs at topic shifts. wording stays the speaker's. nothing is invented.
- guest bio template with name, current role, and one-line context filled from the conversation, plus blank slots for links (twitter, substack, book, company). drop in the rest before publish.
- key quote pulls — three to six standout lines surfaced from the transcript, each rendered as a blockquote with a timestamp link back to the audio:
> "the model is the moat." — guest, [00:21:48]. - markdown and html exports from the same source. markdown for static-site generators (hugo, eleventy, astro), substack, and ghost. html for wordpress, transistor, buzzsprout, and any host that takes pasted markup.
sample output
a 45-minute interview episode produces a show-notes.md file that looks roughly like this:
# episode 47 — naming things, with jane doe
jane doe is a staff engineer at acme corp and the author of
"the naming book." we talk about why naming is the hardest part
of programming, the difference between a good name and a precise
one, and what to do when your team can't agree.
## chapters
- [00:00:00] cold open
- [00:01:34] introducing jane
- [00:04:12] the framing problem
- [00:11:08] precise vs. evocative names
- [00:21:48] when the team disagrees
- [00:34:02] tooling and lint rules
- [00:42:15] what jane is reading
## guest
**jane doe** — staff engineer, acme corp. author of
"the naming book" (2025).
- twitter: [add]
- newsletter: [add]
- book: [add]
## quotes
> "a good name is the one you stop thinking about."
> — jane, [[00:08:30]](#t=00:08:30)
> "the model is the moat."
> — jane, [[00:21:48]](#t=00:21:48)
> "lint rules are an apology for a missing conversation."
> — jane, [[00:36:01]](#t=00:36:01)
html export carries the same structure with <h2>, <ul>, <blockquote>, and anchor links to the published audio's timestamp player.
the workflow that produces it
- upload the episode audio. mp3, m4a, wav — whatever your DAW exports.
- transcript runs. a 45-minute episode is ready in about two minutes.
- AI-tighten paragraphs. one pass that removes filler and breaks runs of speech into paragraphs at topic shifts. you see a side-by-side diff and accept or reject per paragraph. wording isn't paraphrased — only filler and disfluencies are removed.
- editor adds chapter markers. click any sentence to drop a chapter marker at that timestamp; type the chapter title inline. the editor shows the running episode outline as you go.
- mark pull-quotes. highlight any sentence and tag it as a quote. the timestamp anchor is captured automatically.
- export show-notes.md or show-notes.html. paste straight into substack, ghost, transistor, your wordpress block editor, or your static site repo.
when to use this vs. other exports
- show notes export (this page) — for the published episode page, the email newsletter, and the podcast host's description field. the document your listeners actually read.
- clean-quotes export — for journalism workflows that need publication-ready quotes with timestamp footnotes. heavier on verification, lighter on narrative structure.
- SRT / VTT captions — for the video version of the episode going to youtube or instagram. different file, same source transcript.
- blog post draft — when an interview becomes a long-form written piece instead of (or in addition to) a published episode.
privacy
for shows under embargo, NDA-bound guest interviews, or internal corporate podcasts, run the file in private mode. the show-notes export works identically; the audio and transcript stay on your laptop.
pricing for show-notes export
$0.25 per minute, all exports included — show notes, captions, clean transcript, pull-quotes, the lot. no per-format upcharge, no subscription, no minimum. waitlist signups get the first month free and 50% off forever.