Agentic success has a prerequisite — building the systems most enterprises left undone

Enterprises racing toward assured autonomy in agentic AI are running headlong into a decades-old problem: Most of their infrastructure was never designed to be connected — let alone governed at the speed AI demands.

Only about 30% of the enterprise is truly connected today — a fragmentation driven by accumulated technical debt, siloed IT stacks and legacy mainframe systems built across eras that were never designed to speak to each other, according to Dan McAllister (pictured, right), senior vice president for global alliances and channels at Boomi LP. Bridging that gap is now a prerequisite for the agentic enterprise Boomi is building toward — and paradoxically, the very technology creating that urgency may be the only thing capable of meeting it.

“The reason why mainframe still exists is there’s still a lot of value in those systems,” McAllister said. “With the advent of AI, we’re going to see two things: A continued explosion of new applications for technology, and then an increase in the need to go all the way back to the original data sources in the mainframe. Now, that integration — how are we even going to stay at 30%? Well, we’re leveraging AI.”

McAllister and Terese Pate (left), global chief technology officer for enterprise transformation platforms at DXC Technology Co., spoke with theCUBE’s Gemma Allen at Boomi World 2026, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the Boomi-DXC partnership, the barriers holding enterprises back from full connectivity and the path to assured autonomy in agentic deployments. (* Disclosure below.)

Assured autonomy and the trust gap in agentic AI

The connective tissue problem runs deeper than legacy code. DXC Technology — which built a dedicated Boomi practice after nearly nine months of deepened partnership — has seen firsthand how the long-standing divide between engineering teams and the boardroom has been just as much a barrier to full connectivity as technical debt, according to Pate.

“IT was so separate from business — you had the IT stacks and they’re working on it, and then the business had their piece,” Pate said. “This age of AI really brings them together. You still get the business you’ve got to bring to the IT parts … and learning how to speak the same language, to build it, to gain the trust — it takes time.”

That trust question sits at the heart of the assured autonomy challenge. Boomi is addressing it directly through tools such as Boomi Companion — a collection of open-source agent skills for building production-ready enterprise software through natural language — and its intent to acquire Lunar.dev, an AI and Model Context Protocol gateway provider designed to enforce policy-driven control and observability over every agent interaction at scale. These developments all serve Boomi’s wider goal of getting enterprises across the trust threshold, according to McAllister.

“Do I actually trust [agents] to move beyond the four walls of my cube, and can I deploy it across my business?” he said. “In some use cases, yes. Some life-and-death use cases — absolutely not. We have to have the right data feeding it and monitor it and govern it and secure it in the process to ultimately get to a point where, ‘Yeah, we actually do trust the outcome here,’ and we’re in this learning phase now.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Boomi World 2026:

(* Disclosure: TheCUBE is a paid media partner for Boomi World 2026. Sponsors of theCUBE’s event coverage do not have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE