Informatica adds MCP support, spate of AI-fueled features

New capabilities

As AI development continues to surge, much of the current focus is on building agents.

Agentic AI applications -- unlike generative AI (GenAI) chatbots that were among the most widely built tools after OpenAI's November 2022 launch of ChatGPT significantly advanced GenAI technology -- are proactive, rather than reactive.

While chatbots can simplify data exploration and analysis by enabling users to engage with data in natural language, they require users to initiate interactions. Agents possess contextual awareness and reasoning capabilities that enable them to act autonomously, proactively exploring data to surface insights, collaborating to arrive at decisions and taking on tasks to relieve workers of repetitive work.

However, because agents reduce and sometimes eliminate human involvement when making decisions and taking actions, it is imperative that agents are trained to act safely and responsibly.

Model Context Protocol (MCP) is an open standard developed by GenAI vendor Anthropic. It simplifies agent development by defining how they connect with the resources that train them, such as large language models (LLMs) and proprietary data, and how they interact with one another.

AWS, Google and Oracle are among the many data management and AI development providers that already support MCP. Informatica is doing so as well, in a move that the company's chief product officer, Krish Vitaldevara, called "a game changer" for the vendor's users.

"Adding MCP support … means customers can deploy and secure these integrations on their own terms, dramatically reducing setup and integration time from weeks to just hours," he said.

With Informatica's support for MCP, users can build and manage MCP servers to connect to their IDMC deployments, allowing agents and LLMs to access proprietary data and improve agent accuracy and relevance.

Given that MCP is quickly becoming a broadly accepted open standard for enabling agentic AI, Informatica's support for the protocol is necessary to keep pace with other vendors, according to Michael Ni, an analyst at Constellation Research.

MCP is becoming table stakes for data management vendors because enterprises need a reliable, standards-based way to expose trusted data assets to AI agents.

"MCP is becoming table stakes for data management vendors because enterprises need a reliable, standards-based way to expose trusted data assets to AI agents," he said. "Wrapping governed data -- especially master data -- in MCP … bridges the gap between model execution and enterprise trust, dramatically improving the accuracy and reliability of agent-driven decisions."

Aslett likewise noted the value of adding MCP support.

"We anticipate [MCP] will quickly become widely adopted by all data and analytics software providers, giving developers and enterprises confidence that agents and agentic applications can interoperate with their data management and governance tools to take advantage of trusted enterprise data and business workflows," he said.

Beyond MCP support, Informatica's Summer 2025 Release includes the following:

• Claire Copilot for Data Integration, a natural language-based tool that enables users to develop data pipelines.
• Connectors to AI development platforms such as Databricks Mosaic AI, Nvidia NIM and Snowflake Cortex AI.
• An AI governance feature that tracks models, data sets and approvals to ensure responsible and compliant use of AI.
• AI-powered data lineage that automatically tracks the flow of data from source systems to AI models.
• Data Quality Rules as API, a feature that executes real-time data quality checks through REST APIs.
• Elastic Runtime Orchestration, a feature that automatically adjusts the amount of compute power used when executing large data engineering workloads on AWS.
• Automated analysis, including explainability for transparency, of how records are matched so self-service users can hone matching rules without IT support.
• Enrichment and Validation Orchestrator, a tool that provides an AI-powered framework within the Informatica's MDM capabilities to automate validating and enriching records.
• An integration between Informatica's Cloud Data Governance and Catalog and its MDM capabilities that catalogs MDM assets, maps data lineage and enforces security and compliance requirements.
• Connectivity to Delta tables in Microsoft Fabric.
• Recipe Marketplace, an environment for users to share reusable workflow templates.

All but Elastic Runtime Orchestration, which is in preview, are generally available.

Collectively, the platform update includes features that will help users advance their AI development programs, according to Ni.

"This is a strong and strategic release -- not flashy, but meaningful for enterprises serious about scaling governed AI and agentic data workflows," he said. "Informatica's integration of MCP [combined with] elastic orchestration, real-time quality APIs and explainability tools, and you get an architecture that's built for throughput and, increasingly important, trust."

Regarding the impetus for including these specific features in the Summer 2025 release, Vitaldevara cited customer feedback as the main motivation.

Looking ahead

Just as adding MCP support aims to make it easier for Informatica users to develop and manage agentic AI applications, one of the vendor's main focal points over the remainder of 2025 will be on better enabling customers to build AI tools, according to Vitaldevara.

"We remain focused on simplifying AI adoption by establishing a trusted and streamlined path to AI-ready data," he said.

Continuing to focus on enabling AI development, both through product development and integrations with partners, is wise for Informatica, according to Aslett.

"We anticipate that Informatica will continue to roll out Claire capabilities across the IDMC to further enhance automation, while accelerating integration with external AI services to facilitate customer development of agentic applications," he said.

Ni, meanwhile, suggested that Informatica expand beyond providing governed pipelines that deliver trusted data to analytics and AI tools and add capabilities that create a complete decision intelligence infrastructure.

"To meet the rising demands of enterprises, Informatica must extend its reach into unstructured data, semantic search and vector pipelines while offering integrated experimentation and MLOps orchestration," he said. "Data scientists and business users alike now expect end-to-end activation, not just governance."

Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than 25 years of experience. He covers analytics and data management.