Informatica from Salesforce has made a significant headless play with the announcement of new integrations with Google Cloud, Snowflake, and Databricks.
Confirmed at Informatica World 2026 in Las Vegas, the headless premise is that data quality, governance, and master data capabilities should not sit behind a UI, but operate as callable services that AI agents can invoke mid-workflow, on any platform.
The issue that Informatica is attempting to solve here is based on the belief that AI agents handling service interactions have historically been let down by the data beneath them, such as duplicate records, unverified addresses, and stale customer profiles.
Headless data management, as Informatica frames it, addresses that at the architecture level.
Of the three announcements, it is the Google Cloud partnership that carries the most direct CX relevance.
Informatica’s CLAIRE GPT – its conversational AI assistant for enterprise data management – is now generally available on Google Cloud Points of Delivery, enabling data teams to discover assets, assess quality, and resolve governance issues through natural language.
What previously required multi-step workflows can be handled through a conversational prompt.
In discussing the news, Rik Tamm-Daniels, VP of Ecosystem and Technology Alliances at Informatica, said that the announcements “reflect our shared commitment to making enterprise AI more accessible, trustworthy, and interoperable.
“With CLAIRE GPT native to Google Cloud and A2A protocol support, Informatica agents become collaborative participants in the broader Agentic Enterprise — reinforcing our role as the data foundation for AI and our commitment to headless data management, wherever customers choose to build.”
Breaking Down the Walls Between AI Agents
The headline capability in the Google Cloud announcement, though, is support for Agent-to-Agent (A2A) protocol, an open interoperability standard developed by Google. With A2A,
Informatica’s CLAIRE data management agents can collaborate with AI agents across any platform or vendor.
Organizations building on Gemini Enterprise will be able to invoke CLAIRE agents within their workflows to access governed enterprise data without custom integration. That removes a significant barrier for teams trying to build multi-agent customer service architectures at scale. A2A support is planned for the Fall 2026 release.
“The future of enterprise AI relies on open, interconnected ecosystems where agents can seamlessly collaborate using trusted data,” said Satish Thomas, VP of Applied AI and Platform Ecosystem at Google Cloud.
“Our deepened partnership with Informatica will empower customers to build multi-agent workflows with Gemini Enterprise that are rooted in enterprise-grade data management.”
The Same Bet Across Snowflake and Databricks
The headless approach also extends to Snowflake and Databricks.
In both cases, Informatica is integrating its Intelligent Data Management Cloud (IDMC) natively with each platform’s agent-building environment – Snowflake Cortex AI and Databricks Agent Bricks – through Model Context Protocol (MCP) servers.
Developers can invoke Informatica microservices, including metadata search and address validation, directly inside agent workflows without bespoke connectors. Both capabilities are currently in Private Preview, with general availability targeted for summer 2026.
Informatica says it is among the first Snowflake partners to deliver headless integration with Cortex AI.
“From one of our ecosystem’s first MCP integrations with Cortex AI to unified governance for open-data, these capabilities enable joint customers to move faster with AI without compromising on governance or openness,” said Chris Child, VP of Product, Data Engineering at Snowflake.
On the Databricks side, Informatica is also bringing golden record publishing into the picture.
Its MDM platform will automatically push trusted golden records – pre-configured for Customer, Supplier, Product, and Location domains – directly into Databricks SQL.
Planned for October 2026, it offers a cleaner version of something CRM practitioners have been chasing for years: a single, verified customer identity flowing into the intelligence layer that drives agent behavior.
“Customers consistently ask us to help them operationalize agents with well governed data that can scale in production,” said Stephen Orban, SVP of Product Ecosystem and Partnerships at Databricks.
“Our deepening partnership with Informatica gives joint customers the high-quality data foundation they need to deploy AI agents reliably and securely across the enterprise.”
