Turn a podcast transcript into content
Click a button on a podcast episode and get one AI output per prompt in your prompt library, saved to an outputs table in Airtable
Version 1
What it does
A button on an episode record in Airtable calls the webhook. The workflow sets ai_status to writing and reads the episode's name, transcript, guest name and bio, the podcast name and the AI model chosen for it.
The episode's prompt formula field holds several prompts separated by "==", each with a prompt id. The workflow splits them and runs each through the chosen OpenAI model with the episode context. Every result becomes a record in your outputs table, linked to the episode and tagged with its prompt id. When all are done, ai_status is set to complete. It pairs with the pre-podcast workflow, which prepares the episode.