February 2026
This release introduces AI-assisted parsing for data transformations and expands the AI Agent with reference data capabilities.
Parse unstructured text with AI
Extract structured attributes from formatted text like addresses, product codes, or log entries using AI-generated parsing patterns.
The new Pattern parser step lets you:
-
Provide example text data for AI to analyze.
-
Review and refine AI-generated segments and patterns in a guided workflow.
-
Output each segment as a separate attribute for use in downstream steps.
For details, see Pattern Parser Step.
Before: Unstructured product codes as a single text attribute
After: Parsed into individual segments as separate attributes
Reference data tools for AI Agent
The AI Agent can now manage reference data tables, expanding its capabilities beyond data quality and catalog management.
The AI Agent includes 19 new tools for reference data that let you:
-
Search for tables and explore their schemas and sample data.
-
Create new tables and modify existing structures.
-
Add, update, or delete records.
-
Establish foreign key relationships between tables.
-
Assign data quality rules to table attributes.
-
Generate table descriptions.
-
Import data from catalog items using automated transformation workflows.
For details about reference data capabilities and use cases, see AI Agent.
Was this page useful?