Data Management Vendors Promote AI Agents

For most enterprises, getting data ready for AI requires data integration, or the ability to unify and normalize data from disparate and distributed enterprise applications. And thanks to a subset of data management firms known for their integration chops, agentic AI is quickly becoming a vital next step in leveraging that data for optimal outcomes.

Informatica (which has been acquired by Salesforce), for instance, has been showcasing since May 2025 a series of AI agents for use with its Intelligent Data Management Cloud (IDMC) platform. The agents are based on a new CLAIRE AI engine in IDMC that works with Informatica’s data integration and master data management (MDM) products.

Still in private preview, the CLAIRE agents furnish a conversational interface to help prepare data for use in AI applications. An AI Agent Engineering service and associated AI Agent Hub allow users to build their own agents using recipes tagged to a range of popular enterprise databases and services.

An Agentic Data Bandwagon

Informatica’s not alone. Other data integration players are embracing agentic AI. Following is a partial list:

Boomi. Like Informatica, Boomi offers integration platform-as-a-service (iPaaS) solutions. Its Boomi Enterprise Automation platform integrates data from a variety of sources, including applications from AWS, Oracle, Salesforce, and ServiceNow, among others. In addition to its own embedded AI agents, Boomi’s platform supports an Agentstudio and the Model Context Protocol (MCP) for creating and managing custom agents.

Cloudera. This vendor isn’t an iPaaS provider. It offers instead a lakehouse-based “unified data fabric” that secures and governs data via metadata and gives customers access and management capabilities for multiple data stores in hybrid cloud environments. A Cloudera Agent Studio offers no-code and full-code development for AI agents.

MuleSoft. Now owned by Salesforce, MuleSoft is deployed to integrate data from databases and applications that Salesforce doesn’t support directly. MuleSoft also offers a MuleSoft Agent Fabric to collect, govern, and secure multiple agents from disparate sources, including Salesforce Agentforce. The vendor supports the MCP and Agent2Agent (A2A) protocols.

SnapLogic. Using intelligent pre-built connectors called Snaps, the vendor consolidates applications, databases, and APIs. SnapLogic AI is a toolset that includes an AgentCreator that gives users access to popular generative AI models; a SnapGPT copilot; and an AI Agent Showcase that provides access to smart agents for specific applications, such as Revenue Reconciliation and Customer Intelligence agents.

Workato. The vendor combines iPaaS capabilities (supporting 1,200 prebuilt connectors) with Agent Studio, a solution for creating “Genies,” or agents customized to fit enterprise data requirements. The platform supports MCP, with A2A support on the roadmap.

Why Enterprises Welcome Agentic AI

Enterprises seem ready for this next step from data integration to agentic AI. Not only do agents help organize and analyze data, they also allow users to garner specific types of information—about finances, sales figures, and customer characteristics, for instance.

Wescom Financial, for example, a large credit union headquartered in Pasadena, California, has used Microsoft’s and Informatica’s solutions to consolidate data on over 250,000 members in 30 branch offices and to streamline call center operations with OpenAI models. Desigan Reddi, VP of IT and Operations at Wescom Financial, is quoted in Informatica’s press release:

“Informatica’s new AI Agent Engineering service is a game-changer for organizations like ours, enabling us to build and orchestrate intelligent AI agent workflows securely and at scale—without the need for complex coding. The ability to connect agents across our hybrid ecosystem, leveraging trusted data, empowers both our technical and business teams to accelerate automation and drive real-time, data-driven decisions.”

Futuriom will publish a report on “Data Management for AI Pipelines” in January 2026.