Boomi is arguing that the biggest obstacle to enterprise AI adoption isn’t the models themselves, but whether enterprise data is actually usable in real time. At Boomi World 2026 in Chicago, the company positioned “data in motion” as the foundation for what it calls the agentic enterprise, while announcing acquisitions and partnerships aimed at governing how AI agents interact with enterprise systems.
Despite Boomi’s strong focus on agentic AI throughout the event, Chairman and CEO Steve Lucas stressed that AI itself is not the end goal.
“Data is not the new oil, or if your energy source is the sun, data is not the new solar,” he said during his opening keynote. “Data is simply sand. There’s tons of it. Not all of it’s useful, but we can do amazing things with it. We can turn sand into chips and silicon. We can do these things with data, but it’s only if you put it to work. It’s only if you activate that data.”
Citing a study from the Harvard Business Review, Lucas said only 7% of enterprises reported that their data is “in motion,” or is ready for AI. Meanwhile, Gartner said 60% of AI projects will be abandoned this year due to lack of AI-ready data.
“Activating your data is important in this era of AI. Activating your data means getting it out of those hidden caches, the stores of information in devices, systems, and databases, and putting it to work. AI will rapidly consume all of it,” he said.
Lunar heights
As part of its agentic AI push, Boomi announced the acquisition of Tel Aviv-based start-up Lunar.dev, which focuses on AI and model context protocol (MCP) gateways. The proposed acquisition is expected to add governance and management functions for AI usage across enterprise systems into the Boomi Enterprise Platform and Boomi Connect.
“We’ve been partnering with Lunar.dev for some months now,” said Ed Macosky, Chief Product and Technology Officer at Boomi, during a media conference. “They bring capabilities around MCP gateways, governance, and MCP registries, which tie into our Boomi Connect product and other parts of the enterprise platform. We have been expanding our API management capabilities into AI-related workloads, while the Lunar team brings security expertise that we expect to incorporate into the platform regardless of acquisition timing.”
Mani Gill, Senior Vice President of Product at Boomi, elaborated on how Lunar.dev fits into Boomi’s AI strategy. According to him, as organisations connect AI agents into multiple enterprise systems, they need greater oversight and auditing controls.
“Let’s say I am a sales rep, and I use Gong (a revenue AI platform) to record my calls and create transcripts. I then ask AI to summarise all my calls for the day, pull out action items, and update Salesforce. To do that, I need access to two systems, and MCP with connectors helps manage how that access is governed and controlled for the user accessing those systems,” he explained.
Gill said tasks like these require gateways that can provide auditability and governance, although he noted that AI agent traffic presents different challenges from traditional API gateway models.
“We have lots of inbound agents and APIs coming in. Call it API or MCP — that’s the protocol, so we need to rethink how we look at these gateways,” he said.
Gill added that Lunar.dev’s approach centres on micro gateways tied to individual users and workloads.
“By providing micro gateways for each client and user, someone connecting ChatGPT to Salesforce and Gong can have a dedicated gateway with its own policies, auditing controls, and visibility into which tools that user is authorised to access,” he said.
Another area Boomi expects to address through the Lunar.dev acquisition is deciding when workloads should go through AI models versus existing systems. Macosky said this is intended to reduce unnecessary AI processing by directing workloads toward deterministic processes where appropriate.
“That doesn’t always have to involve AI. Right now, many companies are layering AI into every step of a process, consuming tokens and compute resources as AI tries to figure things out, even for tasks like payroll. If an agent needs to run payroll, the intelligent router can send that task directly to the existing payroll process instead of routing it through multiple models first. That’s a powerful part of the story, because that’s what enterprises are struggling with,” he said.
Gill further explained that Boomi separates its control plane, where users design and monitor workflows, from the runtime environment itself. According to him, the runtime can be deployed across on-premises, cloud, or hybrid environments, allowing organisations to position it closer to where their data resides. He added that once agents are placed within that autonomous runtime, the system can determine whether a task should be handled through AI models or existing workflows.
The Lunar.dev acquisition is expected to close within 30 to 45 days, Lucas said.
Strengthened partnerships
Aside from the Lunar.dev acquisition, Boomi also announced several partnerships during Boomi World 2026, including a collaboration with Red Hat focused on agentic AI deployments.
Specifically, Boomi Agentstudio is intended to connect AI agents directly to live enterprise data across applications, systems, and business processes, rather than relying on demo data or sample workflows.
Meanwhile, Boomi Agent Control Tower and Boomi Gateway are designed to provide policy enforcement and governance controls. Boomi’s orchestration layer manages how agents execute tasks and consume compute resources, while Red Hat AI provides monitoring and governance capabilities through its open source AI stack.
Red Hat AI includes a Kubernetes-native runtime for inferencing, governance, and deployment across hybrid cloud environments, including sovereign data centres.
Lucas said the partnership will allow organisations to run models of their choice alongside the Boomi runtime in environments located close to where enterprise data resides. He added that the platform is intended to support different deployment models and customer-selected AI models over time.
Likewise, Boomi is co-engineering an AI infrastructure for agent-based workflows with Couchbase, combining Boomi’s connectivity, runtime, and governance capabilities for AI agents with Couchbase’s vector database capabilities.
Boomi and ServiceNow are also expanding integrations for ServiceNow Workflow Data Fabric, allowing organisations to connect enterprise systems and external data sources to ServiceNow so teams and AI agents can work with real-time enterprise data.
Hybrid infrastructure provider Lightedge, a joint ServiceNow and Boomi customer, is using this combined approach to modernise its CRM and integration landscape.
By combining Boomi and ServiceNow, Lightedge replaced multiple legacy integration tools and consolidated its architecture into a unified platform. According to the companies, this helped streamline workflows, improve data visibility, and support its CRM modernisation efforts involving AI agents and workflow automation.
“Lightedge is modernising its CRM, and the combination of ServiceNow and Boomi supported that transition,” said Michael Gallagher, CIO, Lightedge. “By consolidating multiple integration tools into a unified platform, we reduced complexity and connected data more directly into the workflows used by the business.”
Boomi also announced knowledge platform Guru as a launch partner for its managed connector service, Boomi Connect. The integration combines Guru’s knowledge agents with Agentstudio and the Boomi MCP Registry to make knowledge content available to workers and AI agents across connected systems.
Additionally, Boomi has partnered with revenue platform Gong, bringing revenue signals captured in Gong into the Boomi Enterprise Platform.
Macosky said the partnership is intended to bring Gong’s revenue insights into Boomi’s platform, allowing enterprises to use customer interaction data within automated workflows and Boomi’s automation and AI environment.
Connected enterprise
According to Steve Lucas, Boomi has spent the past 20 years focused on what it calls the connected enterprise, which he described as an organisation that can respond more effectively to operational changes and information across systems.
“You cannot have an AI-driven enterprise, an agentic enterprise, or any other autonomous enterprise unless you have a connected enterprise,” he said.
Lucas said the next stage after the connected enterprise is the automated enterprise, which involves moving data through pipelines, feeds, and real-time streams so information can support workloads and reach people, business systems, and AI systems as needed.
He added that the agentic enterprise builds on top of that existing infrastructure rather than replacing it.
“The agentic enterprise won’t be standalone. It won’t be an island. It will be merged with the infrastructure that enterprises already use today. The agentic enterprise cannot exist without the connected enterprise. Automation will continue to exist as deterministic processes, but future automation will increasingly include agentic capabilities,” Lucas said.
Lucas also reiterated that AI alone is not the defining factor for the next phase of enterprise systems, arguing instead that data moving across systems in real time will be a key requirement.
“Our job is to help you move data, get information to the right place, to the right people, and to the right agents,” he said.
