Boomi launches AI agent control plane for enterprises

Boomi has launched an Agent Control Plane for enterprise AI, aimed at companies seeking tighter oversight of AI agents.

The system sits between AI agents and core business applications to govern actions, manage access and track model spending. It is designed to work across public cloud, private cloud and on-premises environments.

The launch comes as companies move AI agents from pilot projects into live operations, shifting concerns from model outputs to the operational and financial impact of automated actions. Businesses want to know which systems an agent can access, what changes it can make, how much those actions cost and whether decisions can be audited later.

Boomi is positioning the product as a control layer for a fragmented market where organisations often use multiple models, tools and software platforms. Its approach is vendor- and model-neutral, allowing customers to govern agents built on Boomi's own platform as well as third-party and open-source frameworks.

Boomi highlighted functions including visibility into agent and tool activity, policy enforcement through an AI gateway, token usage limits and human approval for higher-risk actions. The platform can also expose business systems as governed tools for AI agents, using existing identity systems to control access and record usage.

This is especially relevant in large organisations, where agents may need to interact with systems such as Salesforce, SAP, Oracle and Workday. Those connections can make AI more useful in areas such as finance, customer service and operations, but they also raise the risk of unauthorised changes, data leakage and compliance failures.

Boomi said the product also addresses data sovereignty concerns by allowing customers to run agents, tools and models within private networks or regional boundaries. That setup can help firms keep sensitive intellectual property behind corporate firewalls while managing compute costs.

Governance gap

Analysts and industry groups have been highlighting the same issues. Gartner has said 40 per cent of enterprises will demote or decommission autonomous AI agents by 2027 because governance gaps often become apparent only after production incidents. The FinOps Foundation reported that 98 per cent of practitioners in its latest survey now manage AI spending, up from 31 per cent two years earlier.

Boomi cited a separate study by Forrester Consulting that found 86 per cent of leaders said their organisations had moved beyond AI agent pilots, but only 34 per cent said they trusted the actions those systems take. According to the research, organisations that deployed agents before they were ready reported an average of USD $2.1 million in added cost.

Michael Barnes, Chief Analyst, Enterprise IT, Omdia, said the market's focus is now on what agents do inside critical systems rather than what they say in a chat window.

"The governance conversation has shifted from model risk to execution risk. Enterprises are no longer primarily worried about what an agent says. They are worried about what it does to a general ledger or a customer record, what that action cost and whether anyone can reconstruct the decision six months later in an audit. Cloud and model providers each govern their own estate, which leaves the average enterprise holding several partial audit trails and no complete one. The organisations getting this right are putting enforcement in the path between the agent and the system of record, rather than alongside it," Barnes said.

Broader platform

The new control plane is part of a broader update to Boomi's enterprise platform. Customers can convert natural language requests into workflows spanning multiple systems, with both a business view and a technical view to help non-technical and IT teams review a process before it goes live.

Boomi is also expanding application programming interfaces and developer tools tied to agent development, data management and platform operations. It pointed to agent trust scoring as one of the newer functions being added around governance and oversight.

Boomi said it already serves more than 30,000 customers and processes more than 200 billion messages each quarter. The company added that its platform holds more than 16 security certifications used by regulated industries and regional markets.

Steve Lucas, Chairman and Chief Executive Officer of Boomi, linked the product launch to growing scrutiny from finance teams over AI expenditure.

"Two years ago, 31 per cent of FinOps practitioners were managing AI spend. This year it is 98 per cent. Finance has arrived in the middle of every AI conversation, and it is asking three questions most architectures cannot answer: which agent spent this, on what, and who approved it. Until an enterprise can answer those questions, AI is an expense nobody can defend, and what cannot be defended does not get funded. That is what Boomi's Agent Control Plane changes: agents visible, actions governed, dollars attributable, running wherever the business requires it to run. Control is not the brake on enterprise AI. It is what makes scale possible," Lucas said.