Informatica has launched a bold new chapter in enterprise data management, introducing autonomous AI Agent Engineering that promises to redefine how businesses build, connect and manage intelligent AI workflows.
Announced during this year's Informatica World event, the centrepiece of the launch is CLAIRE Agents – a suite of autonomous digital assistants designed to automate and optimise the full spectrum of data management tasks goals.
Accompanying this is AI Agent Engineering, a new service within Informatica's Intelligent Data Management Cloud platform, that empowers organisations to build, connect and manage intelligent multi-agent AI systems and compose business applications quickly, securely and at scale.
CLAIRE Agents represent what Informatica describes as "the next evolution in autonomous data management." Unlike traditional automation tools that perform static, rule-based tasks, these agents can reason and make decisions dynamically, based on enterprise-wide data.
Gaurav Pathak, Vice President of Product Management, AI and Metadata used the metaphor of autonomous driving to explain the shift, during a recent interview with TechDay.
"Traditional automation is like cruise control – it keeps things going at a steady pace. Agents, on the other hand, are like a self-driving car," he said. "They plan, adapt to changing conditions and navigate complex environments based on goals, not just tasks."
CLAIRE Agents include specialised assistants such as the Data Quality Agent, Data Lineage Agent and ELT Agent, which are capable of monitoring, remediating and optimising data across complex hybrid ecosystems.
These features are powered by Informatica's Intelligent Data Management Cloud's metadata system of intelligence, a context-rich engine that combines human-curated and AI-generated metadata to ground the agents in the specific needs of each organisation.
"Without metadata, agents are flying blind," Pathak explained. "It's the map of the world for our AI – it grounds them in the unique semantics and structures of an organisation's data landscape."
The company also unveiled AI Agent Engineering, a platform designed to let customers build and connect their own agents across cloud and on-premises environments. This service aims to address growing enterprise demand to create domain-specific AI tools that can work collaboratively and access trusted data across systems.
Sumeet Kumar Agrawal, VP of Product Management, who leads the new service, added that many customers are now looking to evolve from static business processes to agentic solutions that adapt and scale.
"We're seeing a proliferation of agents – every app vendor has their own agents, for example, Salesforce has its own, SAP has its own, etc. – but what's missing is the connective tissue," he explained. "AI Agent Engineering provides the framework to build, connect and manage these agents holistically, so they can solve real end-to-end business problems." Agrawal added that data is the backbone of this vision. "The reasoning of any agent is only as good as the data it has access to. We provide a clean, trusted data foundation so agents can act with confidence," he said.
Both executives stressed the importance of no-code interfaces in democratising AI adoption across technical and non-technical teams. "Writing code is just 20% of the job – maintaining it, securing it and ensuring performance is the real challenge," said Pathak. "No-code makes AI explainable and manageable for everyone."
CLAIRE Copilot, which enables users to generate complex data pipelines using natural language, is now generally available. First launched in preview earlier this year, it acts as a pair programmer for data engineers and complements the agentic approach by giving users interactive control over tasks while agents handle broader goals autonomously.
Informatica's latest strategy also includes broad ecosystem integration, with support for leading cloud and AI platforms including AWS, Azure, Google Cloud, Snowflake and Databricks.
This flexibility, the company says, ensures enterprises can use their preferred AI models while maintaining control over data security and compliance. Security remains a key concern for enterprises experimenting with generative AI. Informatica says its metadata system enforces access controls and data governance rules at every step. "Agents only access data a user is permitted to see, and we've put strict guardrails in place – for example, they can't issue delete commands," said Pathak.
Agrawal added: "Every agent deployment comes with built-in security policies – rate limiting, IP restrictions, authentication protocols – everything an enterprise needs to operate safely."
The announcement has already drawn attention from key Informatica clients. Desigan Reddi, VP IT and Operations at Wescom Financial, described the agent engineering service as "a game-changer". "It enables us to build and orchestrate intelligent workflows securely and at scale – without the need for complex coding," he said. "This no-code, metadata-aware approach aligns perfectly with our vision of making advanced AI accessible and actionable."
As to what the future holds, Informatica's vision is for CLAIRE to become the "front end of data" across the enterprise.
"We want users to simply tell CLAIRE what they want – a report, a pipeline, a governance task – and have the agents take care of the rest," said Pathak.
Asked whether there's such a thing as too many AI agents, Pathak said, "It's not the number that matters – it's whether they're connected and working together to solve the problem."