Orchestrating the Future of Cloud-Native Infrastructure with Shaun O’Meara and Alex Freedland – KubeCon Europe 2025

This is Textron tv. Hey everyone, it's Alan Shimmel, and we're back here at Cube Con. It is a buzzing, buzzing floor.

It's very loud. You know, I was trying to figure out why it looked, not grungy, but, but different than most show floors, and I figured it out. There's no carpeting, there's no, you, you see me from here up, so you can't see down there, but there's no carpeting and it gives it a look.

I, I guess it saved money too. But anyway, look, we're here with one of, one of the, I don't wanna say founding companies of cloud native, but one of the companies that's really put the, put the native in cloud, right? Mm-hmm.

Mortis, if you have, if you're not familiar with mortis, I'm gonna ask this gentleman to give you a quick history. But first, let me introduce them to you. In my immediate left, Sean, Sean Nomar.

Sean is the CTO of Martis, and to my far left Alex Friedland. Alex is the CEO of Martis. Yes.

And the co-founder. And The co-founder. Yeah.

There you go. Alex. I put the pressure on you first.

Let's start it off. Yes. I think most of our audience has heard of Mortis, but maybe some people haven't, and maybe even those who have, are not really familiar.

That's right. Give us the, a quick mortis history. Absolutely.

Well, we are the, the original cloud native company, right? We started in the days of OpenStack when OpenStack became the thing when, um, people understood that, uh, public clouds and AWS is there to stay and got actually threatened by it. And, uh, went together and built a open source standard, um, that became a cloud native operating system.

And we were smack in the middle. We co-founded the open, uh, stack foundation. We became the guys, you know, who kind of built open stack in the early days.

I served on, um, on the board of the foundation for many, many years. And that's kind of how cloud native became cloud native. Mm-hmm.

Uh, then we did it the second time through a different motion because we bought, uh, Dockers Yes. Um, enterprise business. And we became, uh, and still are probably one of the principal, um, uh, managers and, uh, operators of, uh, containers.

Right. And we have probably more than a thousand customers who use our container solution. And now we're kind of in the verge of the new revolution, probably bigger than all of them combined, which is ai.

Right? Right. And of course, AI is not just about applications and agents, it's about how it all comes together in the cloud native way.

So that's a bit Of a history. That's a quick history, but there's a lot wrapped in there, my friend. That's right.

Yeah. And of course, for those who've been following along at home, we actually interviewed yesterday the, uh, the chairman of the infra, was it Open Infra? Open Infra Foundation, which is now merging with Linux Foundation mm-hmm.

And the open in Jonathan, right? Yes. And the Open Infra Foundation is the OpenStack foundation.

That's right. That you're talking about. That's right.

And we've been chatting with him. I said, congratulations of having an exit before I had my hands. And he's like, in our world, exits are not tied money.

I said, well, Still, congratulations. Congratulations. But you know what?

But the important thing was is that the whole OpenStack thing, you know, I, my last conversation, what's old is new again, technology just goes round and round. Right. Cycle.

Yeah. So, OpenStack for a while, you didn't hear a lot about it, and then all of a sudden, I, you know, Rackspace got back involved with it about a year ago. Yep.

Mm-hmm. And it became very big again. And, and now here they are, it's open infra, they've got more than OpenStack.

Correct. They got a secure container, they got a bunch of other stuff, and they're merging into lf. There'll be a, I call it top level domain, a top level foundation within lf.

Absolutely. Amazing. But Morant has had an amazing story.

Right. And I've co-founded four or five companies myself. It's, it's not easy to stay current and ride a wave and hop off one wave and ride the next wave and continue.

Right. It's God bless you. But, uh, thank You.

But let's talk, you know, I, I think that one of the keys for meranti has been, you've, you've seen these standard bodies, standardization and standard bodies, and recognize that aligning your business with the standards is the way to ride the wave. Right? That's, that's how you play.

Correct. And it's not just, it's not just the way, I mean, ultimately we're all in business because we partner with our customers Yeah. To provide a great service to them.

And, um, when you first start a company, it's about a particular technology and a particular expertise and a spot ability to help in something that emerges. But when you graduate and become a true longstanding vendor, it's about partnering with your customers to take them through technology transitions and allow them to use technology as a way to support their business. And for that, you must understand how technology evolves.

You have to stay current, but also you have to stay relevant. And understanding the trends and standards and actually participating in defining those standards is first and foremost what the customers expect you to do. So, um, what's happening now, I believe, is that the reason those foundations came together, because we are at the precipice of the largest transformation related to ai, right?

Uh, it started in the kind of training world which proved, um, you know, the, the, the certain technologies. But it's all being done because you want to actually run AI native applications. Right.

And the, the, the buzzword for that is inferencing. And inferencing is gonna be the largest workload that a humankind has ever seen. And not only is gonna be a workload created by humans, it's gonna be a workload created by ai.

Right. And the more energy, the more power you have, as Nvidia suggested two weeks ago, which was incredible, like 600 watts per rack, all of that is going to be used as intelligence to create new inferencing workloads. Absolutely.

Right. And, and more And more, we're gonna need more than that. It's just a start.

Right. And so you have that as a foundation. You got to have a standard of set of technologies that the industry will coalesce around to kind of move this forward.

And the reason I think this is so exciting for us cloud people, because this type of workload is actually the first real killer app for a hybrid cloud. The one that we never found. Right?

Right. That's exactly right. So we, the cloud people, uh, are beyond ecstatic for ai.

Absolutely. You know, I did, I do a lot of video. I did a YouTube short about a month and a half.

Well, I did a YouTube and the AI cut it into a short mm-hmm. About Mark Cuban said that the world's first trillionaire is not Elon Musk or Jeff Bezos or any of those Bill Gates. None of those.

The first trillionaire is gonna be the guy who creates the killer app for ai. Yes. Using AI in a way we have it correct.

We haven't thought about it. Yeah. And it doesn't have to be human, right.

It could be, it could be an Agent, probably machine. Yeah. But that'll create a trillion dollars of wealth.

Yes. Um, But yet as we sit here today, and and I agree, Alex, I agree with you. I, I'm old.

I remember when the internet became commercial and it, you know, the cloud is cool, but we, people like us get off on the cloud. My wife and her family. Uh, the clouds where they store pictures.

That's exactly Right. The internet changed their life. Yes.

The cell phone we have in our pockets changed their life. Absolutely. AI will change their life Even more.

It's gonna Make the machine accessible to Normal. Yeah. But you can't do it without us, the plumbing people.

Well, right. You still need us for today. Not wood.

But here's the thing that scares me. I remember when the internet had no rules in many ways. There are no rules with ai.

Right. It was a year and a half ago, a hundred scientists or whatever signed the thing. We gotta go slow.

I knew we weren't gonna go slow. Nobody's gonna go slow, slow. Do we ever go slow?

But all of a sudden that's been thrown out the window anyway. 'cause the gold rush is on. Yes.

You know, and who gets rich, who gets rich in the gold rush? People Who take the risks, pick and shovels, Pick and shovel. That's why he's the CEO and you are the CTO, But we need somebody who's smart.

You need, you need the smart people to use the fix and shovels. But it's the people who sell the picks and shovels who make the money. 'cause every sucker in the world is out there looking for gold, but only a small amount of them find gold.

Right. Fair point. But the guy who sells the picks and shovel, he makes money on every one of them.

Wow. Is that what mortis job here is? Well, in the way Yes.

Plumbing is, um, at pix and shovel. Yes. Yes.

And, um, um, we, um, we of course, first and foremost the infrastructure people, right. And infrastructure is everything that goes from, you know, bare metal all the way to the applications. Right.

So it, um, in the, in the modern speak, it's both the infra and the platform. Uh, but today it's also about future proofing and, um, composability, because it's such a wild west, you need to be able to actually, um, curate the ecosystem and deliver the best of breed. And thank God, today with Kubernetes, we're able to, um, not only orchestrate the infra to wherever compute and storage and all of this is needed and available, and it needs to be distributed because of all the sovereign data requirements and the like.

But also because it's an HPC workload and your latency and the speed and usability of your resources is essential. Right. So Kubernetes is now able to do all of that, including, um, orchestrating the platform pieces around and providing the governance for those And the governance.

Exactly. The governance Is very important. It's, it's gonna have to be, you know, I did an interview earlier today with another company And the gist of what they were saying, and this is a company that's DevOps platforms, but they're creating an ML platform, or they created, actually they bought an ML platform.

They integrated. 'cause the, the, the thought was data scientists and the AI app developers of today and tomorrow can't. What we have today in regular Kubernetes and CICD platforms and so forth, is not optimized for that job.

We need a different platform, different base and different platform. Yeah. A different interface.

It DA different set of tools for data scientists, for people who are going to, you know, for AI app developers for the whole inference thing. Yep. Right.

Absolutely. You need a different stack almost. Absolutely.

So is it Kubernetes that we can somehow adapt? Is there something numer is gonna come up with? Uh, um, I've, I've been part of this debate with a lot of techies who say that maybe one day Kubernetes is not gonna be the right tech for it.

And, um, you know, there will be, you know, people are saying, oh, rust will have to come up because of, you know, it's, that's A language. Well, it's a language, but it's actually, uh, much more efficient. Right.

Um, it's like an old, uh, c plus plus versus Java. I mean, it's kind of coming back in the next Same thing again. Right.

But I think as an orchestrator, uh, of components for the foreseeable future, because of the standards and deep adoption, I don't see, I mean, everything that we see is Kubernetes based. There's nothing new on the horizon. Ultimately, Kubernetes is not in the data path.

Right. So Kubernetes runs everywhere. We can leverage Kubernetes to provide for an infrastructure abstraction.

Correct. And that's the key idea behind what we're talking about. At the end of the day, we are gonna use Rust.

We are gonna write applications in rust. We're gonna use more modern languages. Hell, the machine learning models is probably gonna start writing the code that runs the machine learning models.

We're already seeing evidence of that. Absolutely. But we still need control points for all of this infrastructure, and that's ultimately where we're focusing our efforts today.

The plumbing. And the most importantly, and this is again, where we've been having, uh, a lot of debates, I do believe that the Uber orchestration control point has to be completely open source. Yes.

Right. And it's not, I mean, there, there's different types of open source. There is open source, which is open core.

There's commercial open source that now Red Hat has become the masters of, And Hashi And Hashi and all that. And There's all IBM by the way. Well, yes.

All IBM. Um, And, and then there is kind of the, the, the kind of real open source, uh, like, like OpenStack was Right, right. When the, the contributors themselves are in foundations and they actually develop it out, the Open, it's free in freedom, not free is in beer.

Exactly. And, um, I think for this layer, because it's such an important control point for everything that happens, the ultimate thing needs to be through open source. And, um, we've built a new project, uh, which is based on our K Zeros, um, um, ecosystem, which since I came back to s last year, um, we've contributed this to CNCF, our K Zeros and Coron projects.

But we've built, um, a product, um, called, I don't even call it a product, I call it an environment, right. Called Coordinate. With the zero that we built from day one, we built it, um, as Apache two in GitHub.

Really. And, um, we are of course, you know, major contributors to it today, but we already have people from other companies contributing. And it's open source first and True open source, true open Source and good when time is right.

Um, so we built it is really is an underpinning for the world of inferencing. Right. It's, it's also a good thing for a world without inferencing.

And we have our existing customers that do traditional computing that love that because it's something that can orchestrate clusters on AWS and on Google and on OpenStack and VMware and what have you. Right. And that's, that's a big use case in traditional computing, but ultimately it's done for inference.

And I think it's only a matter of time because the market forces that created OpenStack is ultimately OpenStack was an answer to a need in the market that came as a result of people needing to have a cloud native operating system as an answer to AWS. Right. Kubernetes came for the same thing at the, um, at the application level right now, AI native, I mean, that's why the two foundations came together.

So they can have one voice and one power to actually focus on ai. There has to be an open stack of ai, however, however it's gonna be called, however it's gonna be named. Um, but an ecosystem needs to come together and people need to understand that they need, they need to speak with one voice, because only one voice will bring the marketing power and awareness.

Right. And standards. And I don't know if it's gonna happen in three months or in a year.

It's gonna happen. It's gonna happen. I mean, Those standards also bring safety for the consumers of those systems.

It brings Stability Vendor out of the right Throat To Choke as they Future proofing and all that. So the Coordin product that we've built environment to us is something that we want to, uh, one day become put out there for that, for that standard Kubernetes and, And to prove that we have this need. Right.

We've actually announced the launch of it in January this year. So it's been like two months, Three months. Three months.

Yeah. We have a catalog of people who are embracing it and writing about it and all that. I think it's what 30 entities now moving.

Fantastic. Uh, And it's like five come in every day and it's not something that we are pushing them to do. I know.

It's, it's there, It's organic because, Because the need is there and people No, I, I agree. Let me, let me throw something else out at you. And I'm, I have my own thoughts on this.

So everyone is saying ag agentic ai mm-hmm. Agentic AI is gonna be the next big thing here. It is.

Yeah. Are we gonna have one super agent that does multiple tasks or we have a lot of ephemeral agents that do one thing and they, it's gonna Be a lot of ephemeral agents that do very specific tasks. Do we need a Kubernetes for that?

Absolutely. And is that this Kubernetes or is this some sort of super Kubernetes for a agent? AI Agent AI models at the end of the day, are just another workload that runs within the Kubernetes ecosystem.

They have special requirements, they have special needs, they have underlays for, you know, access to specialized infrastructure. But ultimately what we need is a way to distribute those workloads where the resources are. And that's the thing.

We live in a resource constrained world. You know, you might be spending $50 million in GPUs from Nvidia today, but when those are used up, they're used up, where else are you gonna find these resources? And this is where Kubernetes comes into the ecosystem.

It allows us to create these multiple points of access, access, distributed all these resources, distribute these workloads, and then distribute the application layer on top of that. And, and you can't have it all on Amazon because your, that little agent needs to run at, let's say, where, where data, data is either sold or generated. Right.

So if you, it's Gotta be on the edge. It's gotta be everywhere. It's Be everywhere.

I mean, we have, in our booth, we have a demo where we have edge devices right there. The Nvidia, what is it thing? Jetson Nano jet.

Yeah. Yeah. The $250 or something.

Yeah. Yeah. And it's, it's, it's, you access it through 5G uh, hotspot.

Familiar. Right. And you just discover it, right.

And it becomes part of your availability. And the model just goes there and it's cashed through some cashing mechanism. So it's instantaneous.

And I mean, that's what the world is gonna look like. And of course, AWS is gonna be a huge part of it. And so is Google All that?

The all will be, it's Everywhere. It's gonna be one of so many. Right.

And that's why it has, it has to be that open source fabric that is, It's the only you'll Get, it's beholden to nobody To do this. Right. You need coopetition.

Yes. And to do that, you need a foundation. You need True source.

Yes. That control point should be beholden to nobody. Absolutely.

That's the only way. And and that's something Morant's action stands for as well. Mm-hmm.

Yeah. We're, we're probably over time. I, I get it.

They're giving me tan signals here already, but for people who wanna stay up on this, yeah. Obviously go to Ntti. But where, where, where would you else, would you send them GitHub Com thetis Slash Cordon with a zero.

Okay. Can you spell the whole Word? K zero R-D-E-N-T.

Thank You, Sean. io and if you want to see the catalog partners or you want to join the catalog, program catalog, do accordant io. I love it.

Sean, thanks for coming on here, Talia. Appreciate it. Alex, this was great.

Thank you so much. Thank you. Check out Meranti, we, we got people you can't see 'cause they're on the other side.

They're lined up for lunch. But, um, thank you for sitting in on this. We'll have more on this as the days and weeks unfold, but we're here at Q Con.

This is Alan Shimel. We're out.