Reel scripts and captions from a session

A button on a session record in Airtable calls the webhook. The workflow uses the session transcript, or downloads its VTT caption file and turns it into a timestamped transcript if there is no transcript yet.

For each reel prompt selected on the session, it fetches the prompt's instructions, creates one placeholder record per script in your reel series table marked AI Writing, then asks a model for that many scripts with captions, using your brand context, event name, speaker and transcript. Each placeholder is filled with its script and caption and set to To Do. If the model fails, the placeholders are set to Error. The reel scheduling workflow posts these once the final video is uploaded.