AI Field
An AI Field allows users to automatically generate API-ready data by providing natural language instructions during the workflow setup.
Common Use Cases
Use the AI Field when the workflow requires complex data that is difficult to construct manually:
Filter Generators: Creating complex JSON filter objects (e.g., for Notion or Pipedrive).
Query Builders: Generating dynamic search conditions or database queries.
Payload Construction: Converting simple user prompts into formatted API JSON payloads.
Metadata/Content: Creating structured content blocks or dynamic metadata.
How an AI Field works?
Step | What Happens |
|---|---|
1 | User enters instructions |
2 | Optional schema/data is sent using |
3 | AI reads the prompt and available context |
4 | AI generates API-ready output |
5 | The generated output is stored and later used in the workflow |
Configuration
Add Field → Select AI Field
Use the following JSON structure to define an AI field in your action or trigger:
Property | Value | Example |
|---|---|---|
Prompt | Instructions that tell the AI what kind of output to generate and in what format | Generate valid Notion filter JSON using the provided database schema |
suggestionGenerator | Provides additional field information or sample data to help the AI generate more accurate output | Return database column names and field types |
[...][ { "key": "filter_conditions", "type": "aifield", "label": "Filter Conditions", "help": "Generate valid Notion filter JSON using the provided database schema", "required": true, "prompt": "Write a prompt to guide the AI.", "suggestionGenerator": " // Enter JS code response of which will be sent to AI later used by ai to generate results" } ]
Additionally, attributes like visibilityCondition can be used as per need and requirement.
