Boomi World 2026 | Steve Lucas, Boomi

Clips
More from Boomi World 2026

Steve Lucas

CEO

Boomi

play_circle_outline From Proof to Profit: Deploying AI Agents with Measurable Enterprise ROI and Disciplined Investment Planning

play_circle_outline Within Two Years: Governing and Orchestrating Evolving Agent Skills, Integrating ERP/CRM for an Agent-Run Enterprise

play_circle_outline Domain-Specific AI in Regulated Industries: Compliance, Observability, Governance and Automation Outcomes (Lexitas, Multiquip)

play_circle_outline AI builder persona blends developers and domain experts; shadow AI presents risks.

play_circle_outline Boomi's Lunar Acquisition Enhances Runtime Containerization, Enabling Private On‑Prem Model Deployment with Red Hat OpenShift

Info
Transcript

Steve Lucas, Boomi

Steve Lucas

CEO Boomi

In this interview from Boomi World 2026, Steve Lucas, chairman and chief executive officer of Boomi, joins theCUBE's John Furrier and theCUBE + NYSE Wired's Gemma Allen to discuss the evolution of intelligent integration from back-office infrastructure into the orchestration and governance layer for enterprise AI. Lucas draws a sharp distinction from prior years: AI agents are no longer a preview — they are deployed and generating real ROI. He frames Boomi as the single harness across data, applications, APIs and AI models that organizations need to operation... Read more

explore Keep Exploring

Why did the speaker wait six minutes before saying "AI" in the keynote, and what point were they making about enterprise AI adoption and driving ROI? add

Will AI agents soon perform the majority of enterprise work, and how should organizations govern and orchestrate these agents as they develop and acquire skills? add

How should organizations implement and govern AI—particularly in regulated industries—to achieve ROI while ensuring compliance and explainability? add

How do you envision the "AI builder" (a combined developer and domain expert) persona evolving in enterprises, especially given the rise of shadow AI and the need for closer collaboration between business and engineering? add

What impact do observability and containerization have on the evaluation and operational deployment of AI agents, and how can organizations harness multiple models while keeping their data private? add

bolt Powered by CUBE AI

Steve Lucas, Boomi

search

Gemma Allen

>> Hello, and welcome to theCUBE, coming to you from Boomi World 2026, here in the windy city that is Chicago. I'm Gemma Allen, joined by my co-host, John Furrier, and we are just fresh off what was an excellent keynote. Joining us now is the CEO of Boomi, Steve Lucas. Welcome, Steve.

Steve Lucas

>> Thank you. It is awesome to be here. I'm thrilled you're here as well.

Gemma Allen

>> Well, that was a great keynote, right? You covered a lot of my favorite things, my father's favorite movie, Trading Places, my kids' favorite thing, cereal. You got a lot of very relative, I think, and relatable analogies in to message what has been a very interesting time for you and this company. But maybe let's start with what has changed, Steve, one year on from the last time you were here? We were at Boomi World in 2025.

Steve Lucas

>> Last year, we were talking about AI agents, and honestly, the year before that as well. This year the agents aren't coming, they're here. AI is starting to become not just pervasive conversation in the enterprise, it's a pervasive, and not even experiment, it is being deployed. It is real. We're starting to see real ROI coming from AI and agents, and enterprise processes, integrations, automations. It feels very real now, but organizations are still asking questions like, "Well, okay, so we're starting to experiment. We're seeing some early ROI, but what comes next? If we're going to put all these dollars into AI, how do we really drive ROI?"

John Furrier

>> I loved the keynote, because I was trying to get it out of you when you came to the NYSCQ Wired Studio. You said "Stay tuned," but you kind of did give some teasers about containers on the runtime side, which was awesome.

Steve Lucas

>> Yep.

John Furrier

>> But I have to ask you, you went six minutes before you said AI, and you kind of said the two letters. Was that a set piece, or was that like an internal bet, over/under, will you come clean and say it early? Because you held suspense there, because you were obviously making fun at the hype, but also it is the top of the keynote.

Steve Lucas

>> Right. Well, it is. It is the topic. I was having a little fun with it, because I think every organization rushes in, and we all do. But part of it is there's an analogy there, which is with AI we tend to rush in. And now that coding is easy, I guess, and we can do all these things with AI, we forget the parts like planning, thinking ahead of time. What do we want to do? So it was kind of intentional. I think it worked. The point is that we're moving so quickly with AI that we don't stop and think things through, which by the way, can be detrimental, because if you get started with AI too quickly, you don't bring people along, you don't do the analysis, then you end up failing.

John Furrier

>> Well, I'm glad we have some time here with you to unpack some of the things we have questions on. But I like how you held the tension in the room, because you were making the point that I took away, was there's a lot at stake. The timing is right. You had great reference to Newton, laws of physics, mentioned some Steve Jobs references. But you kind of teased it out, like, "Look, this is a moment in time for opportunity, yet it's the same game in the enterprise complexity." Share your thoughts on that, because you guys had announced the big news, and talked about it on stage, the Lunar acquisition.

Steve Lucas

>> Yeah.

John Furrier

>> That's a huge piece to the runtime aspect. Talk about how you see that business impact going on, what does that all mean?

Steve Lucas

>> Well, look, every large software company, technology company out there thinks they have AI figured out. They've got a plan for everybody. They'll tell you exactly how you can harness AI. But here's the truth, we're still very early innings with AI, and we're starting to see the enterprise landscape settle into a couple key things. We know building agents is a clear functional part of the enterprise, so is governing agents, because things like trust and transparency, that matters in the enterprise, so we're starting to see that play out. Other organizations are coming in with things like agentic gateways, and the like. But I feel like we're still so early, and still trying to understand, well, what risks do probability engines, let's call them what they are, really pose to organizations, especially in the enterprise? And my point is that AI as transformative as it is, as capable it is... I heard Dario, I mentioned this on stage during my keynote, talk a couple days ago about how AI is reaching near human-level intelligence. Well, that's terrific, but we still have to figure out how all this stuff functions inside of a much broader organization. By the way, the enterprise is subject to silly little things like regulations and rules and risk management, so we have all that. So, knowing that, what we wanted to lay out for people was we absolutely believe that AI will be the transformative technology, we know that, but that landscape of data, applications, APIs, processes, systems, and now AI models and frameworks, all of these things that you have to think about, there has to be a single way to harness that capability. I talked specifically about that, it's how do you harness all those things to leverage AI for good? That's the case that we laid out, is that Boomi can be that harness for your enterprise.

John Furrier

>> It was a very clever keynote. I liked how you weaved in, first of all, the messaging was great, but you kind of weaved on, not a poke, but the elephant in the room is the SaaS apocalypse.

Steve Lucas

>> Yes.

John Furrier

>> And I like how you talked about the referee and rules, and you said "ERP and CRM are the rules."

Steve Lucas

>> That's right.

John Furrier

>> Everyone's talking about, "Oh, vibe coding's going to kill Salesforce. CRM's dead. ERP is dead." So, you kind of made a point intentionally, that those systems just get buried, my words, not yours, into the system. This speaks to the governance piece. Explain your thinking and vision around that, because I think that's really unique, because system of records have been around for a long, long time, but now we have a system of execution with agents.

Steve Lucas

>> Yeah.

John Furrier

>> So okay, tie that together, why are ERP and CRMs either not going to die, and if they do die, why would they die?

Steve Lucas

>> Well, first of all, I think the point you're getting at, which is really critical, I believe with 100% certainty that in two years AI agents will do the vast majority, not all of it, but the vast majority of work in the enterprise that we do today. But as humans are creative, we'll find new work to do. We always do. It's like my wife keeps saying to me, it's like, "Don't you ever just sit still?" No, I find new work. We're going to find new work, that's what we do. And so AI, the agents that we have, they'll do more work, so that's going to happen. But this governance word that you're getting at. Governance is kind of a funny word, because governing means yes, applying rules. So, don't access this data. Even if someone's asking for it, don't do it anyway, right? So, governing AI is only part of it. It's also orchestrating AI. How do we take AI agents, keep an eye on them as they evolve, but also understand their skills. Think about it, if you hire somebody out of college, they're going to learn. They develop, they grow. As we get older, we get more skills. AI agents will gain more skills. So for us, governing is also about understanding AI and its skills as it develops in your enterprise. So, we're inventorying those skills, and then we're enabling organizations to put those agents to work. So if I said, for example, "I want to create an employee onboarding process," I can summon agents that have the skills to onboard employees, like one agent to get your laptop, on agent to help you figure out where all the PowerPoints are, one agent to help you figure out how do I run expense reports inside of my business. Those agents will get marshaled by our orchestration technology. So yes, the point on governance is not just watching agents, but understanding them as they grow. We are the parents.

Gemma Allen

>> And Steve, when we think about governance and trust, especially from the perspective of some of the long-tail opportunities in this era, like sovereign AI, AI on the edge, governments across the world who really need to have a position whereby they are in control of what is going to become a hugely important asset for the future of the citizens of any one country. What sort of cases and use cases are you seeing there? What sorts of conversations are you having with partners and clients that you weren't having maybe two years ago, and how urgent do you think is that conversation?

Steve Lucas

>> Well, I think it's incredibly useful. So here's what we know, on the left hand, people are panicked in the boardroom. Executives at companies want to have people that come in, and say, "Don't worry. I know everything about AI. I've got it all figured out." And again, as I said in my keynote, if someone says that, then they've got something to sell you. What we need to do right now, is we need to come together and figure out, first and foremost, what are the operating parameters for that organization? Maybe AI can automate all my capability, but I'm a bank, I'm a government institution, I'm a healthcare organization. HIPAA doesn't go away just because you have AI. You can't say, "Well, that was a bad decision, but AI made it." It still has to be explainable. So, the first thing that organizations want is domain-specific AI, not generalized AI, that generally helps me. If I'm a healthcare organization, I want HIPAA-compliant, healthcare-specific, domain-specific AI. They want that, so that's number one. And then number two, they want to understand, well, how does this plug into my business process? So, what I talked about on stage today, Lexitas is a case in point. Lexitas has automated almost 50%, it's 46, but almost 50% of their payments which are highly-regulated, processed today, or prior to AI, processed manually, now they're using AI agents to process those with zero human involvement. Now you'd think, "Well, that still needs oversight," but actually the oversight is happening through the governance plane with Boomi. So, that's happening today. I talked about another one, Multiquip. That's another great case in point. They had their technicians that would have to sort through thousands of manuals for the machinery that they manufacture when they're dealing with support issues. Almost 80% of that is happening through AI, through agents on the Boomi platform. So the ROI is there, but we're going to go from generalized ROI to domain-specific ROI very quickly.

John Furrier

>> I have to ask you on that, one of the things we've been fleshing out on theCUBE is this persona called the AI builder, but it's not a developer-only, it's a developer and a domain expert.

Steve Lucas

>> Has to be.

John Furrier

>> What is your vision on this persona, domain builder? Because there's a lot of shadow AI going on, and it's not the same shadow IT we saw. Shadow IT was actually innovative. Go around your boss, put your credit card down, go to the cloud, get a prototype, get your hand slapped, and then you get promoted.

Steve Lucas

>> Right.

John Furrier

>> That's cloud. Now we've got AI happening, shadow AI happening in finance, all the domains because of the tools.

Steve Lucas

>> Indeed.

John Furrier

>> So, how do you see that persona evolving? And like DevOps, which were developers operating the cloud, you're going to see AI builders operating their own AI. What is your vision on that? Because you cross both personas, developer and business integration.

Steve Lucas

>> I think we're going to see more collaboration between business and development than we've ever seen before. Engineering has traditionally been over here, business functions over here, never the twain shall meet. And yet here we are. Part of what's driving that, and this is a brilliant point that you bring up, is it's almost too easy to build software now, to the point where if everybody's the founder and CEO of their own software, or whatever it is, which I even poke fun at myself. Because look, every single week, I get down a Claude code rabbit hole, I'm building something that I think is helpful, and I take it to my CEO, and he goes, "Oh, my gosh, Steve, thank you." But there's all these other things that you need to think through in the enterprise. What would have been better is collaborating with that business partner, and so we have to see that. So, I think organizations are going to evolve. Traditional org charts, where engineers live over here, subject matter experts live over there. AI is deleting those rules. Those are going away.

John Furrier

>> Yeah. We've been talking to a lot of CFOs, because they're involved in a lot of these transformations. To your point, vibe coding, you mentioned this on stage with your example, is not just... It's got to be integrated, which you guys do that. CFOs are closing the books the last day of the month at midnight, and that wasn't vibe coding, that vibe code a prototype integrated with the rules of the road, AKA their systems of record, and they're done. That, to me, is at builder. Do you agree?

Steve Lucas

>> I agree completely. I agree. That scenario is playing out everywhere right now, but there's risk associated with it as well, because two things are happening. One is the acceleration of innovation. Look, 95% of the stuff that gets built right now with AI is still not well thought through. There's going to be way more software before there's less of it, and we all know that, but that's just on the right hand. On the left hand, what organizations need to do right now, is they need to get, A, their arms around their data. Put AI aside, you've got to get your arms around your data. Where is it? Is it of quality? Can I deliver it in real time? Because here's the real risk, if I'm struggling, if you're struggling to deliver data to human intelligence in real time with quality, how do you think that's going to play out for AI that operates thousands of times faster.

John Furrier

>> Yeah. Garbage in, garbage out.

Gemma Allen

>> When we think about that, about moving from a data world that we've known for 15, 20 years now, to this world of data activation and liquid data, right, and some of the challenges that have existed historically that have been there for a long time, in terms of how batch data, how data moves, how rigid it is, how governed it is. It seems as though the narrative is that we've had some sort of magic response to this, right? That overnight, this problem can be solved. What are your thoughts on some of the challenges that really still do remain at enterprise level, especially for highly regulated industries like some of the ones that Boomi work in? And how far away do you think we are from having this almost fluid, seamless, liquid data solution across the board? How real is it versus the narrative?

Steve Lucas

>> I think that we have a ways to go. We're getting there. Most organizations, especially if you're a CIO, have heard pitches like a data fabric, a data mesh. I love liquid data, I'm totally going to borrow that. But they've heard those pitches, but the reality when the rubber meets the road is some systems, like mainframes, which still exist in the vast majority, especially in financial institutions, insurance organizations, those systems alone, they're not necessarily real time. They're difficult to integrate with. Or you have legacy applications, which by the way, the majority, I'd say 60, 70% of companies out there still have massive legacy ERP systems that they're still struggling with. These are not liquid data systems. And even with the amazing advancements that we have with Snowflake, Databricks, and the like, you can read from those systems, but can you write back? All of this stuff has to evolve. It's going to take years to make that happen. But our mission, build the connected enterprise, drive the automated enterprise, and enable the agentic enterprise, that's what we're focused on. So while I am going to steal the "Liquid data" term, I think that's where we need to get to. It's going to take years.

John Furrier

>> Yeah. And by the way, those systems you're mentioning, ERPs, they weren't designed to take liquid data.

Steve Lucas

>> No.

John Furrier

>> They were designed to just report.

Steve Lucas

>> Correct.

John Furrier

>> Also, they were built in longer development cycles, huge R&D budgets, massive old-school thinking, but that's what's out there. Now when you look at AI, people have to be AI-ready. So I want to ask you, what's it take to be AI ready? It's funny, I was talking to a friend, Paul Saffold wrote an article on Harvard Business Review in 1997, called Are You Machine-wise. And he was talking about, are you ready for the complex, interconnected, more central conduct of business? That's evergreen from 1997. That was internet. How did that apply to agents and AI? Because people are asking themselves the question, "How do I become AI-ready? Am I AI-wise?" What would you answer that question with?

Steve Lucas

>> Question number one must be, across all of my IT data, as well as my OT data. So if you're in power production, transmission, utilities, manufacturing, you're going to have all this OT and machine data that's latent, dormant, untapped, unleveraged, unmonetizable. We want to wake that data up. Not just what sits in your IT systems, but in your OT systems as well. So step one, wake it up. Find it, make it connected. That's why we talk about the connected enterprise. I said from mainframes to MCP, and everything in between, which there's been a lot of that, we want to find it and we want to build the connected enterprise. Every single system in your enterprise must be connected, number one. Once you've built that connected data, the next set of Cs are not just connected, is it consistent? Is it contextual? These things matter as well. The vast majority of companies, if I said today, "Hey, is your data accessible in real time? Is it connected? Is it consistent? Is it contextual?" For humans, no, it's not. How is it going to work thousands of times faster with AI? So, building the connected enterprise, step one, then step two, ensuring that you have the security framework and the automations to start building agents. So, if you do those things, then your data is going to get active and it's going to get ready for AI.

John Furrier

>> There's a control point too. I want to ask you about observability, because that's translating into the AI world. The container piece is huge you guys have now, in play.

Steve Lucas

>> Yes.

John Furrier

>> How does that impact the operational size of evaluation of agents, the observability piece?

Steve Lucas

>> This is our partnership that we talked a little bit about with Red Hat, but here's a belief that I have. I asked a question on stage, which is "Which AI model is right for your company," right? We're thinking, well, is it Claud? Is it OpenAI? And the answer is all of them. That's the answer. So the next question is, well, if all of the models are the right answer, how do I harness those? How do I containerize them? So that's what we're partnering with Red Hat, is our OpenShift AI partnership with Red Hat, where the Boomi data runtime that processes your connections, your workflows, in concert with whatever model you want. If you want an open weight model where you keep your data privatized, I mean, my gosh, organizations are hemorrhaging data to the cloud, and all of these open-frontier models, how about you keep it private in an open-weight model? So having that container, with our runtime, with those models, now you can privatize AI. You can build your data geometry, your graph that is unique to your company, not help somebody else build their trillion-dollar company. That's what that container's about.

Gemma Allen

>> Well, Steve, I think we could talk to you all day. That was certainly a fantastic keynote, and I think the energy here, at Boomi World, is absolutely palpable, but I'm getting the nod from the production crew. So maybe just to finish out, I know you're at the man at the moment, and it's a busy few days ahead yet, right, what else is on the agenda for you over the next couple of days?

Steve Lucas

>> Well, it's talking to amazing people from theCUBE first and foremost. I love having theCUBE here. It means the world to me. For me, it's actually now we've got a record amount of both press and analysts here at Boomi World. Boomi World has just grown massively, and we're thinking, "My gosh, where do we do this next year? We got to find more room." And partly, that's the great coverage that you two help us provide as well. For us, it's about building the data activation business for every single one of our customers here. That's what we're focused on, harnessing AI for the future, not talking about eliminating jobs, but building them.

Gemma Allen

>> Well, I love to end with that, building jobs and building the future. Steve Lucas, thank you so much for joining us on theCUBE. I'm Gemma Allen, here at Boomi World 2026, joined by my co-host, John Furrier. We have a great lineup of guests coming on the show today, some great conversations. Stay tuned.