Informatica Added AI Agent Data Cataloging
The company has updated its Claire engine to make metadata machine-readable for automated data stewardship tasks.
Updated on Sept. 23, 2026 in Artificial Intelligence
Live Poll
Do you trust artificial intelligence tools to provide accurate data for your important business decisions?
Informatica has updated its AI engine, Claire, to enable machine-readable metadata catalogs for AI agents. This change aims to provide trust indicators for AI-generated answers by linking data directly to verifiable system definitions.
Why it matters
As enterprises deploy AI agents, these systems can no longer rely on human tribal knowledge to interpret corporate data dictionaries. Providing machine-readable metadata allows agents to automatically verify the sources of their outputs, increasing the reliability of data-driven decisions.
Users of the Claire data quality agent can now generate 200 data quality rules per hour, a significant increase from the previous manual rate of 4 rules per week. The system functions by extracting metadata definitions from existing data dictionaries and utilizing a maker-checker process for multi-stage approval.
The players
Informatica
An enterprise data management company owned by Salesforce that provides metadata connectivity for cloud data platforms.
Claire
The proprietary artificial intelligence engine developed by Informatica to automate data stewardship and quality management tasks.
PepsiCo
A multinational food and beverage corporation that utilizes Informatica software to locate revenue data definitions.
Gaurav Pathak
A representative who presented Informatica's latest AI metadata capabilities at Dreamforce 2026.
The details
Informatica processes metadata by ingesting definitions directly from established system data dictionaries rather than relying on manual documentation. The system implements a maker-checker process, where AI-generated definitions must undergo multiple automated approval checks to ensure accuracy. These indicators are then integrated into AI platforms like Claude to verify the provenance of specific figures.
Timeline
Salesforce acquired Informatica for $8 billion in 2025.
The Informatica World conference took place in May 2026.
Gaurav Pathak presented at Dreamforce in San Francisco on September 23, 2026.
The Tech Race
Informatica extends the data governance capabilities established by platforms like Microsoft Purview by making metadata natively machine-readable for agentic AI. This move positions Informatica to capture the growing market for reliable AI data orchestration against established cataloging competitors.
Enterprises using Salesforce and connected cloud platforms can immediately leverage Claire to automate data stewardship, reducing manual rule creation requirements. Organizations will likely see improved accuracy in AI-generated reports as agents gain the ability to trace data back to source dictionaries.
The takeaway
The transition to machine-readable metadata is essential for moving AI agents beyond basic text generation and into reliable enterprise analytics. Stakeholders should monitor future integration updates with Google and Snowflake platforms to see if these trust indicators gain cross-platform standard status.
Further reading
For more information on how current systems handle automated data stewardship, visit the Artificial Intelligence section.
Source note: This article includes information reported by Computer Weekly.
Live Poll
Do you trust artificial intelligence tools to provide accurate data for your important business decisions?
