Skip to main content

​
What This Node Does

The Filter node removes rows that don’t meet your specified conditions, allowing you to focus on specific subsets of your data. It supports multiple conditions with AND/OR logic, date ranges, text matching, and null value handling. [SCREENSHOT: Filter node on canvas showing conditions and row count change]

​
When to Use This Node

Use the Filter node when you need to:
  • Reduce dataset size - Remove irrelevant rows early in workflow for better performance
  • Focus on specific timeframes - Show only recent data (e.g., last 30 days, Q1 2024)
  • Exclude invalid data - Remove null values, test accounts, or incomplete records
  • Create segments - Extract specific subsets (e.g., high-value customers, active users)

​
Step-by-Step Usage Guide

1

Add Filter node to canvas

2

Connect to upstream data

3

Add first condition

Select column, operator (equals, greater than, contains, etc.), and value[SCREENSHOT: Filter condition configuration]
4

Add more conditions (optional)

Toggle between AND and OR logic to control how conditions combine[SCREENSHOT: Multiple conditions with AND/OR toggle]
5

Preview filtered results

​
Tips and Best Practices

Filter Early: Place Filter nodes immediately after Input nodes to reduce data volume before expensive operations like joins or aggregations. This can speed up workflows by 10-100x.
Use Smart Dates: Instead of hardcoding dates like 2024-01-01, use smart values like “last 30 days” so filters automatically update when the workflow runs.
Check Row Counts: After adding a Filter node, verify the row count makes sense. If it’s 0 or unexpectedly low, your conditions may be too restrictive.
Null Handling: For columns that may have missing values, always add an “is not null” condition to avoid unexpected results in calculations.
Case Sensitivity: Text operators like “equals” and “contains” are case-sensitive. “West” does not match “west”. Use Formula nodes with LOWER() or UPPER() functions for case-insensitive matching.
Test Incrementally: Add one condition at a time and preview results. This makes it easier to identify which condition is causing unexpected filtering.