Workflows where every step can be real code
A flow is a chain of steps from top to bottom. Each step is a script, a SQL query, a decision, a loop or a ready-made action, and each one reads what the steps above returned.
Code where you need it, ready steps where you don’t
Write a step in any supported language, or drop in a PostgreSQL or MySQL query. Sending e-mail, calling an HTTP API, transforming a list or reading a file are built-in steps.
Learn more →Conditions, loops and parallel branches
Branch on a value, loop over the rows a query returned, or start named branches together and read each result by its branch name.
Learn more →A person approves before it goes on
An Approval step pauses the flow, holding no worker, until someone approves or rejects. Approved, the same run continues with the values they filled in.
Learn more →Questions
Can I edit a flow as a file?
Yes. View file shows the whole flow as YAML, and a project can keep its flows in a Git repository.
How do I start a flow?
From the editor, on a schedule, through an HTTP endpoint, or from an event such as a new e-mail.
How many rows can a SQL step return?
10,000 by default; the step’s row limit can be raised.