Thanks to everyone who joined for this month's AI Crew session!
Connecting ChatGPT to Airtable
Ryan opened with the basics: how to connect Airtable and ChatGPT, authenticate, and choose which bases to grant access to (specific bases or all of them). Once connected, ChatGPT can see your base schema, records, and comments, giving it real structured context instead of one-off prompts with no grounding.
Reading and reasoning across your base
Using a fictional e-bike company launching in Seattle as the example, Ryan showed ChatGPT reviewing a live campaign, summarizing overall readiness, and citing the exact Airtable records behind each conclusion. The key idea: because your data is relational, ChatGPT can reason across linked records (work streams, partners, approvals, channels) instead of just pulling a single record.
Writing back to Airtable
From there, Ryan showed ChatGPT creating new linked action items tied to risks, connecting them to the right parent tasks and supporting records, and rewriting a risk summary in concise, executive-friendly language while preserving specific human-owned fields exactly as they were.
Building interfaces and decks in minutes
One of the biggest wow moments: Ryan had ChatGPT build a full executive dashboard interface, complete with charts and risk callouts, in under two minutes. He also showed it turning all the work from the session into a presentation deck, with a note that using your own branded template will get you on-brand output instead of the generic default.
Automations and skills
Ryan previewed an upcoming automation feature (rolling out for ChatGPT soon) that builds a weekly digest pulling action items, release notes, and channel performance into an AI-summarized email. He also touched on "skills," reusable instructions so you don't have to repeat formatting or output preferences every time you prompt.
Have you connected ChatGPT + Airtable? Let us know in the comments what you’ve been using it for! 👇
Hope to see you at the next one!
