Airtable automation that enriches rows and asks first
Create, update, find and delete Airtable records from a workflow, enrich new rows with an AI agent, and gate deletes behind an approval. Token connect.
- Category
- Productivity
- Connect
- Paste an API key · managing connections
- Operations
- 4
- Triggers
- none yet
Airtable is where a lot of teams keep the list that runs the business: leads, tickets, content calendars, inventory. MadMax turns that list into something that acts, in both directions. A new row can start a run through the watch trigger, which polls list_records on a schedule and fires once per unseen record (it baselines existing rows first, so a table with 500 historical records does not start 500 runs). An AI agent can read the row and fill in the columns nobody wants to research by hand, or hold any record operation as a tool when the job calls for judgment rather than a fixed sequence. A person can approve the change before it lands back in the table. Describe it in plain English and MadMax builds the workflow on a canvas, runs each step against real records from your base, and shows you the result before anything goes live.
What teams build
- A watch trigger polls the Leads table every 15 minutes, an AI agent with the browser tool reads each new company's domain, and
update_recordwrites size and a one-line summary back to the row with typecast on, so "Enterprise" lands in the single-select field instead of failing. - A hosted form files support requests into the Tickets table with
create_recordafter an AI classify step labels each oneurgent,normalorquestion, and an if branch texts the on-call person for the urgent ones. - A weekly schedule lists every record where
{Status}="Archived", an approval emails the owner the count and a sample of names with one-click approve or reject, and only on approve does a loop calldelete_recordfor each match.
How it connects
Paste a personal access token. Create one in Airtable under Account, Builder Hub, Personal access tokens (it starts with pat) with data.records:read, plus data.records:write for creating, updating and deleting, on the bases you want MadMax to reach. Paste it into Settings → connectors → Airtable or straight from a step's connection picker on the canvas. The token is encrypted at rest and resolved only inside your workspace at run time, and rotating it is a reconnect in place. See managing connections.
Related
- Watch trigger: fire once per new row.
- Approval step
- AI agent step
- Managing connections
- MadMax vs Zapier
Every Airtable operation MadMax can run
Straight from the integration's manifest, so this list is what the steps on the canvas can actually do today.
| Operation | What it does |
|---|---|
| create record | Insert a single record into a table. |
| delete record | Delete a record by ID. |
| list records | List records in a table. Returns the matching page. |
| update record | Update a single record. PATCH only touches the fields you supply; unmentioned fields stay as-is. (Use PUT semantics via execute_query if you need full-replace.) |
Works with
Describe the workflow. MadMax builds it and tests it on your data.
Free to start, no card. Connect your apps, say what should happen in plain English, and watch each step run against real data before anything goes live.