Podcast Show Notes Generator
Example prompt: "I'll drop the transcript for each new episode into a Google Doc named 'Episode XX — Transcript'. When that happens, read the transcript, write show notes that include a 100-word episode summary, 5-7 chapter timestamps, three pull quotes from the guest, and a list of every book, person, and link the guest mentioned. Save the notes to a new doc called 'Episode XX — Show Notes' and post a link in our #podcast Slack channel for the producer to review."
The Problem
Producing show notes is the unglamorous tail of every podcast episode. After the recording and the edit, someone still has to listen back (or re-read a transcript), pull out chapter markers, identify quotable moments for social, list every book and person mentioned, and write a tight summary for the episode page. Most independent shows either skip it, do it inconsistently, or burn an hour per episode on a job that is mostly mechanical extraction.
How GloriaMundo Solves It
We build a scheduled workflow that polls the transcripts folder every hour, picks up any new transcript documents that do not yet have a matching show notes doc, and produces a complete show notes package for each. An integration step pulls the transcript text from Google Docs. An LLM step writes a 100-word summary of the episode. A second LLM step identifies natural chapter breaks and proposes timestamps. A third LLM step picks three strong pull quotes suitable for social. A fourth extracts every named book, person, product, and URL the guest mentioned. A code step assembles all of this into a structured document. An integration step writes the show notes to a new Google Doc and posts a Slack message linking the doc for the producer to review and publish. Glass Box preview shows the full draft before anything is written.
Example Workflow Steps
- Trigger (scheduled): Runs every hour against the transcripts folder.
- Step 1 (integration): List recent transcript docs in the Google Drive folder and pick any that lack a matching show notes doc.
- Step 2 (integration): Read the full transcript text.
- Step 3 (LLM): Write a 100-word episode summary.
- Step 4 (LLM): Propose 5-7 chapter titles and the timestamp at which each begins.
- Step 5 (LLM): Select three pull quotes from the guest, with attribution.
- Step 6 (LLM): Extract every named book, person, product, and URL mentioned, with a one-line description.
- Step 7 (code): Format the parts into a structured document with consistent headings.
- Step 8 (integration): Create the show notes Google Doc and post a Slack link for review.
Integrations Used
- Google Drive — list new transcripts in the polled folder
- Google Docs — read the transcript, write the show notes
- Slack — notify the producer when a new draft is ready
Who This Is For
Independent podcast producers, in-house podcast teams at publishers, and agencies producing client shows where show notes are part of the deliverable but not the most exciting part of the week.
Time & Cost Saved
Producing show notes manually for a 45-minute episode typically takes 60-90 minutes of careful re-reading and copy work. This workflow reduces it to 10-15 minutes of review and final polish per episode. For a weekly show, that is roughly an hour back per week; for a daily show, several hours.