Unifying Cloud-Native Architecture

Hey everyone, we're back here at SUSECON on our Tuesday... No, Wednesday coverage. The day zero thing on Monday had me all messed up.

So we called this day two, and it's really day three. In any event, let me introduce you to Sudeep Goswami. Sudeep, if you watch Techstrong TV a lot, the last time I interviewed you was Atlanta for KubeCon, and that was probably now...

Well, it wasn't a year, it was probably six to eight months ago. Six months. Yeah.

Because we just came out of KubeCon in Amsterdam. Amsterdam. Sudeep is the CEO of Traefik Labs.

But before we get into Traefik Labs, before we talk about SUSE, let's talk, Sudeep, to these people. Give them an idea of how you came to be the CEO here, and your passion for what you're doing. Very simply, first of all, great to be here, Alan, with you again.

Thank you. I love tech. Okay.

And I love solving problems. You put those two things together, and that's what brings me here. Absolutely.

What about Traefik Labs? Yeah. Traefik Labs, a very, very amazing company.

The product has been around for 10 years. Mm-hmm. Developers love it.

I started in the company about three years, so that's when my journey with Traefik started. I came in as the CRO- Yep ... and then became the CEO.

So I've been the CEO for about two years now. Mm-hmm. And it's absolutely a great place to be.

For someone out here who says, "I never heard of them," what do they do? Yeah. So we're set at the intersection of cloud native, Kubernetes native, and now AI native workloads, and our job is governance, and it's runtime governance.

What does that mean? We provide a gateway which allows customers to unify their routing, their security, and their observability. We have a huge open source project that's called Traefik.

Mm-hmm. And then we have the paid tiers, which is Traefik Hub. Okay.

So you talk to any developer, nine times out of 10, chances are they've heard of Traefik, or they're using it in their home lab, or they're using it in their company. So look, the whole governance, the word governance, the whole aspect of it, is really taking on a different shadow in light of AI. Right?

When you have AI generating 41% or more of the code out there, and we're generating 2X, 3X, 4X more code than we were before. When you have AI scanning for security vulnerabilities and finding 2X, 3X, 4X more vulnerabilities. As humans, we just don't have those kinds of resources.

Right. And so governance becomes the governor, right? Governance becomes the next bottleneck.

Exactly. And not a bottleneck, maybe a break- It's a rate limiter ... on a flywheel, on a runaway train.

" Now governance is, I think, code for, we got to keep a human in the loop here, or at least a human at the helm using AI. But governance is that choke point. Governance is that point where, not a roadblock, but where at least we can get a handle on these things.

Right. Because they're scaling beyond our ability to really have a handle on them. Part of that is there's this convergence of three waves, if you will.

Right? I know you've spoken to me about it. We talked about it off camera.

There's modernization... Excuse me. There's migration- Migration ...

modernization- Transformation ... transformation. " Could be.

But there's different aspects to each of them. Share with our audience what you mean when we talk about migration, modernization, transformation. Yeah.

And so let's start with what is driving all of this? Before the whole AI, let's say, revolution started, customers were already dealing with migration and modernization. Yeah.

And what does that mean? In the migration, because of VMware, and the choices that they have made, that has created a forcing function in the industry for people to look for alternatives to ESX or other parts of the VMware stack. Mm-hmm.

So this was a migration wave that had already been going on- Yes ... before AI really accelerated. Right?

It started around 2023. And then you have the modernization, which has been going on for many years as well, and it started with decomposing a monolith, which is a VM, into microservices. And then people soon realize, well, it's not going to be a binary world of either VMs or containers.

It is going to be a coexistence of both for many, many years. So these two initiatives had already been going on for a while, and then two big things, two additional forces came in. On the migration side, Ingress NGINX, which is an open source project, was powering over 50% of the cloud native workloads.

They decided at the last KubeCon that they're going to retire that project. This was six months ago. Yes.

You and I were there. Mm-hmm. And now it has happened.

So at the end of last month, at the end of March-At KubeCon Amsterdam, that project is now retired, and many enterprise customers who have been using this piece of technology are looking for an off-ramp and an on-ramp to something new. And then you have AI, which is the transformation arc. This is where you infuse AI into everything, right?

So what is now happening for enterprises, they don't have the luxury to treat these as three linear initiatives that they can kind of sequence one after the other. No. They're just going to get out-competed in the market, and their peers are going to pretty much- Eat their lunch.

Exactly. So they have to figure out how do they do all three of these at once? I think this is the biggest challenge that we're seeing every customer, every enterprise face.

How do they modernize? Well, first of all, how do they migrate? Because that's day zero job.

Mm-hmm. How do they modernize? That's day one.

And while they're doing this, how do they transform their business infusing AI? Love it. Now, so here's the interesting thing, though.

If you do this right, you can do all three in one motion, right? But part of that is picking the right partners, picking the right solutions, having the right strategy. Talk to me about the relationship with SUSE and how that plays into it for you.

Yeah, you said it right, Alan. It is the choices that a company or a partner or a vendor makes, whether it's SUSE, whether it's us, it has to be very strategic. It has to be thought through.

You have to think about the future as well. And what we arrived at working with SUSE, that the approach comes down to architecture. You create an architecture that is adaptive, is dynamic, it's future-proofed, it can scale.

It is not locked into its sovereign, what we call sovereign smart. Yep. All of these attributes that the architecture needs to have, and this is the way that you can actually solve this in a very thoughtful, strategic manner that gives you operating leverage, because that's what it comes down to, is do you have the operating leverage so every decision that you make compounds in value.

And so let's start with the migration. The architecture that you put in, you lay in place to allow that migration to happen, should give you leverage to then do the modernization and then do the integration of AI with transformation. And this is why I'm excited about what we have done with SUSE, and we can talk more about that.

Well, let's go ahead. All right. So let's talk about the architecture.

What SUSE has done a great job is they have unified the platform. So they call this the Rancher suite. And then you have SUSE Virtualization as one of the key building blocks of that.

Yep. This is for VMs to migrate over. And then you have the Rancher Prime- Yeah ...

which is for the container of the world, the Kubernetes of the world. Right. And where you have, and SUSE is unique in this way because they have an edge play with K3s.

K3s. And they have an enterprise play with RKE2. Right.

And what Traefik is doing, partnering with SUSE, is we become the default ingress layer for all three of those estates. So great news is we had already been partnering with SUSE, with K3s for many, many years. And because of the Ingress NGINX forcing function, if you will, that happened.

SUSE chose to also standardize on Traefik Labs for their RKE2. And now at the upcoming next release of Harvester, which is their SUSE Virtualization, which is coming up in the fall, we are going to be the default on the virtualization as well. So now you can start to see the value- Yeah ...

which compounds in benefits, where you have a unified layer, which is portable. You can deploy this anywhere SUSE can be deployed, which is on-prem, public cloud, private cloud, the edge. And now with us and SUSE, we have a unified governance layer that follows the workload, wherever that workload may be.

One thing I didn't hear you mention was the SUSE AI, though. And I'll- Like transformation. And now...

Exactly. So now with this foundation in place, now let's talk about SUSE AI. And the team has been doing great work.

You heard at the keynote today. Yes. MCP servers are, there's two sides to AI.

There's AI on infrastructure and infrastructure for AI. Right. So, and we have a great role there to play with them.

As SUSE introduces MCP servers to their customers, those MCP servers need to be governed. So make sure that the agents getting access to these MCP servers can only do the things that they're allowed to do. Conversely, with the SUSE AI Factory, you're going to be exposing models and doing inferencing and tokens as a service.

This is where you need runtime governance as well. And you started with the governance. The way we kind of see governance, there's three things there from a runtime perspective.

A great governance solution allows you to be able to write rich policies, so fine-grain policies with the right metadata. Second, it allows you to enforce those policies in real time. And third, it allows you to observe what happened.

Yeah. Did the policies actually work? Actually work.

Was there anything that bypassed the policy? And you can observe it through agentic means, but also human in the loop means. Yeah.

You don't have to pick and choose, or you don't have to choose just one. Choose one or the other. They're not mutually exclusive.

You want to harness the best of both worlds. Use man in the middle when needed, but then also use agentic. Love it.

This is what I'm talking about, is you got to take a very architectural approach to this because then it sets you up for the next decade of innovation. Fantastic. Sudeep, we're about out of time.

15 minutes go so quick here. We never mentioned your website. io.

io. io. Hey, man, I love what you're doing.

Always a pleasure. Keep doing it. I hope I don't have to wait till Salt Lake City to see you again, though.

Yeah, let's do this again soon. You know anytime. All right.

All right. io. We're here at SUSECON.

We're going to be back with more. Stay tuned.