Workflows now include a native AI-powered action that extracts structured data from unstructured text.

AI Data Extract can take raw input such as emails, SMS messages, webhook payloads, or AI outputs, and return clean, typed values that you can use in later workflow steps.

To set it up, define the fields you want to extract, choose the input source, and optionally provide extra context to improve accuracy.

Key capabilities

• Custom value input: Select any text-based data in your workflow as the input source.
• Optional context: Add supplementary context to improve extraction for complex or ambiguous inputs.
• Templates: Start from a pre-built template or create your own field set.
• Downstream ready: Extracted fields are saved as workflow variables for use in later actions.

How to use it

1. In your workflow, click + Add Action.
2. Search for AI Data Extract in the AI action category.
3. Select the input source (for example: email body or webhook payload).
4. Optionally add extra context.
5. Choose a template or define your custom fields.
6. Save. The extracted fields will be available in downstream actions.

Why this matters

This makes it easier to turn unstructured text into structured variables directly inside the workflow builder, without manual workarounds or external tools.