Everyrow icon

Everyrow

AI-powered data operations: rank, dedupe, merge, screen, and agent workflows

Actions7

Data Operation → Screen

AI-generated

Summary

Filter input data rows by applying AI-based criteria defined in a user-provided task description, removing rows that do not meet the criteria.

Inputs

  • Session Name — Optional name for the Everyrow session to group operations, visible in the Everyrow dashboard. Defaults to 'n8n-everyrow-session'.
  • Task Description (required) — Text describing the filtering criteria to screen rows. Rows that do not match this criteria will be excluded.
  • Response Schema (JSON) — JSON Schema defining the expected response format from the AI. Must include at least one boolean field to determine if a row passes the filter (default schema provided).

Output shape

a single record containing task metadata

The node returns an object with task_id, session_id, status, and operation type ("screen"), representing the submitted screening task. The actual filtered data is obtained asynchronously via a separate task result operation.

Examples

Example 1: Screen a list of companies to only include those that have raised Series A or later funding.

Set 'Task Description' to a filtering prompt like "Filter companies that have raised Series A or later funding". Optionally customize the Response Schema or Session Name.

Links

Discussion