 Welcome back to the city that never sleeps. Lisa Martin and John Furrier in New York City for AWS Summit 22, with about 10 to 12,000 of our friends. And we've got two more friends joining us here today. We're going to be talking with Hasi Bujani, one of our alumni, co-founder and CEO of Ruff A Systems, and Kevin Coleman, Senior Manager for GoToMarket for EKS at AWS. Guys, thank you so much for joining us today. Thank you very much for having us. It's great to be here. It's great to be back at an in-person event with 10, 12,000 people. Yes, there are a lot of people here. This is packed. It's a lot of energy here. So, Steve, we've got to start with you. Your t-shirt says it all. Don't hate, K8s. Talk to us about some of the trends from a Kubernetes perspective that you're seeing, and then Kevin will get your follow up. Yeah, absolutely. So I think the biggest trend I'm seeing on the enterprise side is that enterprises are forming platform organizations to make Kubernetes a practice across the enterprise. So it used to be that a BU would say I need Kubernetes. I have some DevOps engineers. Let me just do this myself. And the next one would do the same. And the next one would do the same. And that's not a practical long-term for enterprise. So this is now becoming a consolidated effort, which I think it's great. It speaks to the power of Kubernetes because it's becoming so important to the enterprise. But that also puts a pressure because what the platform team has to solve for now is they have to find this fine line between automation and governance, right? I mean, the developers, they don't really care about governance. I don't, just give me stuff. I need compute. I'm going to go. But then the platform organization has to think about how is this going to play for the enterprise across the board? So that combination of automation and governance is where we're finding, frankly, a lot of success in making enterprise platform team successful. I think that's a really new thing to me. It's something that's changed in the last six months, I would say, in the industry. I don't know, Kevin, if you agree with that or not, but that's what I'm seeing. Yeah, definitely agree with that. We see a ton of customers in EKS who are building these new platforms using Kubernetes. The term that we hear a lot of customers use is standardization. So they've got various ways that they're deploying applications, whether it's on-prem or in the cloud and region, and they're really trying to standardize the way they deploy applications, and Kubernetes is really that compute substrate that they're standardizing on. Kevin, talk about the relationship with RAF-FACE systems that you have and why you're here together, and two, second part of that question, why is EKS kicking ass so much? All right, go ahead. First one, the relationship. It's probably your favorite question. Second one, EKS is doing pretty well. Yep, yep, yep. So yeah, we work closely with RAF-FACE, excuse me. A lot of joint customer wins with Hasib and Co. So they're doing great work with EKS customers and yeah, love the partnership there. In terms of why EKS is doing so well, a number of reasons, I think. Number one, EKS is vanilla upstream open source Kubernetes. So customers want to use that open source technology, that open source Kubernetes, and they come to AWS to get it, in a managed offering, right? Kubernetes isn't the easiest thing to self-manage. And so customers, you know, when back before EKS launched, they were banging down the door at AWS for us to have a managed Kubernetes offering. And, you know, we launched EKS, and it's just, there's been a ton of customer adoption since then. You know, Lisa, when we grew up 12 years and everyone knows we started in 2010, we used to cover a show called OpenStack. I remember that. OpenStack Summit, and at that time, at that time, Kubernetes wasn't there. So theCUBE was present at creation. We've been to every KubeCon, every CNCF, then took it over. So we've been watching it from the beginning. And it reminds me of the same trend we saw with MapReduce and Hadoop. Very big promise. Everyone loved it, but it was hard. Very difficult. And in Hadoop's case, big data, it ended up becoming a data lake. Now you got Spark, a snowflake, and Databricks, and Redshift. Here, Kubernetes has not yet been taken over, but instead, it's being abstracted away and or managed services are emerging. Because general enterprises can't hire enough Kubernetes people. They're not that many out there yet. So there's a training issue, but there's been the rise of managed services. Can you guys comment on what your thoughts are relative to that trend of hard to use, extracting away the complexity, and specifically the managed services? Yeah, absolutely. You want to go? Yeah, absolutely. It's important to not kid ourselves. It is hard. But that doesn't mean it's not practical. When Kubernetes is done well, it's a thing of beauty. I mean, we have enough customers who scale, like just, you know, it's like a, not forget a hockey stick. It's a straight line up because they just are moving so fast when they have the right platform in place. I think the mistake that many of us make, and I've made this mistake when we started this company, was trivializing the platform aspect of Kubernetes. Right, and a lot of my customers, you know, when they start, they kind of feel like, well, this is not that hard. I can bring this up and running. I just need two people. It'll be fine. And it's hard to hire, but then I need two, then I need two more, then I need two more. It's a lot, right? I think the one thing I keep telling, like when I talk to analysts, I say, look, somebody needs to write a book that says, yes, it's hard, but yes, it can be done. And here's how. Let's just be open about what it takes to get there, right? And I mean, you mentioned OpenStack. I think the beauty of Kubernetes is that because it's such an open system, even with the managed offering, companies like Graphic and really productive businesses on top of this Kubernetes platform, because it's an open system, I think that is something that was not true with OpenStack. I've spent time with OpenStack also. I remember how it is. Well, Amazon had a lot to do with stalling the momentum of OpenStack, but your point about difficulty, Hadoop was always difficult to maintain and hiring against. There were no managed services and no one yet saw that value of big data yet. Here at Kubernetes, people are living a problem goal. I'm scaling up. And so it sounds like it's a foundational challenge. The ongoing stuff sounds easier or manageable. Once you have the right tooling. Yeah, no, I mean, once you have the right tooling, it's great. I think, look, I mean, R, I mean, you've never talked about this before. I mean, the thesis behind RAFE, is that, you know, there's like eight, 12 things that need to be done right for Kubernetes to work well, right? And my whole thesis was, I don't want my customer to buy 10, 12, 15 products. I want them to buy one platform, right? And I truly believe that in our market, similar to what vCenter, like VMware's vCenter did for VMs, I want to do that for Kubernetes, right? And the reason why I say that is because, see, vCenter is not about hypervisors, right? vCenter is about hypervisor, access, networking, storage, all of the things, like multi-tenancy, all the things that you need to run in enterprise-grade VM environment. What is that equivalent for the Kubernetes world, right? So what we are doing at RAFE is truly building a vCenter, but for Kubernetes, like a kCenter. I've tried getting the domain, I didn't, I couldn't get it. Well, after the Broadcom deal, you don't know what's going to happen. Yes. I won't go there. Yeah, let's not go without today. Kevin, EKS, I've heard people say to me, love EKS, it just adds serverless, it's a home run. There's been a relationship with EKS on some of the other Amazon tools. Can you comment on what you're seeing as the most popular interactions among the services at AWS? Yeah, and was your comment there add serverless? Add serverless with AKS at the edge, and things are kind of interesting. Well, one of the serverless offerings we have today is actually Fargate. So you can use Fargate, which is our serverless compute offering, or one of our serverless compute offerings with EKS. And so customers love that. Effectively, they get the beauty of EKS and the Kubernetes API, but they don't have to manage nodes. So that's a good amount of adoption with Fargate as well. But then we also have other ways that they can manage their nodes. We have managed node groups as well in addition to self-managed nodes also. So there's a variety of options that customers can use from a compute perspective with EKS, and you'll continue to see us evolve the portfolio as well. Can you share, Hosebe, can you share a customer example, a joint customer example that you think really articulates the value of what RAFE and AWS are doing together? Yeah, absolutely. In fact, we announced a customer very recently on this very show, which is MoneyGram, which is a joint AWS and RAFE customer. Look, we have enough, the thing about these massive customers is that not everybody's going to give us their logo to use, but MoneyGram has been a RAFE plus EKS customer for a very, very long time. At this point, I think we've earned their trust, and they've allowed us to kind of say this publicly, but there's enough of these financial services companies who have standardized on EKS. So it's EKS first, RAFE second. They standardize on EKS, and then they looked around and said, who can help me platform EKS across my enterprise? And we've been very lucky. We have some very large financial services companies, some very large healthcare companies now who, A, EKS, B, RAFE. I'm not just saying that because my friend Kevin's here. It's actually true. Look, EKS is a brilliant platform. It scales so well, right? I mean, people try it out relative to other platforms, and it's just a no-brainer. It just scales. You want to build a big enterprise on the backs of a Kubernetes platform, and I'm not saying that's because I'm biased. Like, EKS is really, really good, right? That there's a reason why so many companies are choosing it over many other options in the market. You're doing a great job of articulating why the theme of the New York City Summit is scale anything. Oh yeah. There you go. Oh yeah. I did not even know that, but I'm speaking the language. You are! One of the things that we're seeing also, I want to get your thoughts on guys, is the app modernization trend. Because unlike other standards that were hard that didn't have any benefit downstream because they were too hard to get to, here Kubernetes is feeding into real app for app developer pressure. They got to get cloud-native apps out. It's fairly new in the mainstream enterprise, and a lot of hyperscalers have experience. So I'm going to ask you guys, what is the key thing that you're enabling with Kubernetes in the cloud-native apps? What is the key value? Yeah. I think there's a bifurcation happening in the market. One is the Kubernetes engine market, which is like EKS, AKS, GK. Back in the day we used to call operations and management. So the OAM layer for Kubernetes is where there's need. People are learning. Because as you said before, the skill isn't there. There's not enough talent available in the market. And that's the opportunity we're seeing. Because to solve for the standardization, the governance and automation that we talked about earlier, you have to solve for, okay, how do I manage my network? How do I manage my service mesh? How do I do chargebacks? What's my policy around actual Kubernetes policies? What's my blueprinting strategy? How do I do add-on management? How do I do pipelines for updates of add-ons? How do I upgrade my clusters? And we're not done yet. There's a longer list, right? This is a lot, right? And this is what happens, right? It's just a lot, right? And really the companies who understand that that plethora of problems that need to be solved and build easy to use solutions that enterprises can consume with the right governance and automation, I think they're going to be very, very successful here. Because this is a trend, right? I mean, this is happening, right? It's not us, it's happening, right? Enterprises are going to keep doing this. And Open Source is a big driver in all this. Absolutely. And I'll tag on to that. I mean, you talked about platform engineering earlier. Part of the point of building these platforms on top of Kubernetes is giving developers an easier way to get applications into the cloud. So building unique developer experiences that really make it easy for you as a software developer to take the code from your laptop, get it out of production as quickly as possible. So is that what you mean? Is that tie your point early about that vertical straight up value once you set up it right? Because it's taking the burden off the developers for stopping their productivity. Absolutely. To go check and is it configured properly? Is the supply chain software going to be there? Who's managing the services? Are we orchestrating the nodes? Yep. Is that automated? Is that where you guys see the value? That's a lot of what we see. In terms of how these companies are building these platforms, it's taking all the component pieces that has he was talking about and really putting it into a cohesive whole. And then you as a software developer, you don't have to worry about configuring all of those things. You don't have to worry about security policy, governance, how your app is going to be exposed to the internet. It sounds like infrastructure is code. Come on. It's what it's here. Infrastructure is code is a big key to it for sure. Infrastructure is code. Infrastructure is sick as code too. Security, huge. Well, it all goes together. I guess we talk about developers on service, right? The way we enable developers on service is by teaching developers, here's a snippet of code that you write and you check it in and your infrastructure will just magically be created. But not automatically, it's going to go through a check, like a check through the platform thing. These are the workflows that if you get them right, developers don't care, right? All developers want is I want compute. But then all these 20 things need to happen in the back. That's what, if you nail it, right? I mean, I keep trying to kind of pitch the company. I don't want to do that today. But if you nail that. I'll give you a point at the end. But I just have a tangent question because he's reminding me. There's two types of developers that have emerged, right? You have the software developer that wants infrastructure as code. I just want to write my code. I don't want to stop. I want to build it, shift left for security, shift right for data, all that's in there. I'm coding away. I love coding. Then you got the under the hood person. Yes. I've been to the engine. Certainly. So that's more of an SRE data engineer. I'm wiring services together. Yeah. A lot of people are like, they don't know who they are yet. Sure, in college or they're transforming from an IT job. They're trying to figure out who they are. So the question is, how do you tell a person that's watching? Like, who am I? Should I be just coding? But I love the tech. What do you guys have any advice there? You know, I don't know if I have any guidance in terms of telling people who they are. I mean, I think about it in terms of the spectrum. And this is what we hear from customers is some customers want to shift as much responsibility onto the software teams to manage their infrastructure as well. And then someone to shift it all the way over to the very centralized model. And we see everything in between as well with our EKS customer base. But yeah, I'm not sure if I have any direct guidance for people. Any wisdom? If you're coding more, you're a coder. If you let them play with the hardware or the gears. Look, I think it's really important for managers to understand that developers, yes, they have a job, they have the right code, right? But they also want to learn new things. It's only fair, right? So what we see is developers want to learn and we enable for them to understand Kubernetes in small pieces, like small steps, right? And that is really, really important because if we completely abstract things away like Kubernetes from them, it's not good for them, right? It's good for their careers also, right? It's good for them to learn these things. This is going to be with us for the next 15, 20 years. Everybody should learn it. But I want to learn it because I want to learn not because this is part of my job and that's the distinction, right? I don't want this to become my job because I want to write my code. If you're more attracted to understanding how automation works and robotics or making things scale, it might be under the hood. Yeah, look under the hood but then in terms of who keeps the lights on for the cluster, for example. All right, see, Kubernetes is a lot of value. Now you know who you are, ask these guys. Congratulations on your success on EKS2. Quick, give a plug for the company. I know you guys are growing. I want to give you a minute to share to the audience a plug that's coming. What are you guys doing? You're hiring, how many employees? Funding, customers are new wins. Take a minute to give a plug. Absolutely, I come see John. I think every show you guys are doing a summit or a coupon, I'm here. And every time we come, we talk about new customers. Platform teams at enterprises seem to love RAFE because it helps them build that, well, Kubernetes platform that we've talked about on the show today. I think many, many large enterprises on the financial service side, healthcare side, digital native side seem to have recognized that running Kubernetes at scale or even starting with Kubernetes in the early days, getting it right with the right standards. That takes time, that takes effort and that's where RAFE is a great partner. You provide a great SaaS offering which you can have up and running very, very quickly. Of course, we love EKS, we work with our friends at AWS but it also works with Azure, we have enough customers in Azure, it also runs in Google, we have enough customers in Google and it runs on-premises with OpenShift or with EKS A, whichever option you want to take. But in terms of that standardization and governance and automation for your developers to move fast, there's no better product in the market right now when it comes to Kubernetes platforms than RAFE. Kevin, while we're here, why don't you plug in for EKS too, come on. Yeah, absolutely, why not? So yes, of course, EKS is AWS's managed Kubernetes offering. It's the largest managed Kubernetes service in the world. We help customers who want to adopt Kubernetes and adopt it wherever they want to run Kubernetes, whether it's in region or whether it's on the edge with EKS A or running Kubernetes on Outpost and the evolving portfolio of EKS services as well. We see customers running extremely high scale Kubernetes clusters, excuse me, and we're here to support them as well. So yeah, that's the biggest- And I'll give the plug for the cube. We'll be at KubeCon in Detroit this year. Lisa, look, we're giving a plug in everybody, come on. We're plugging everybody. Well, speaking of plugs, I think, Haseeb, you have a book to write, I think, on Kubernetes, and I think you're wearing the title. Well, I do have a book to write, but I'm one of those people who does everything at the very end, so I'll never get it right. So if you want to work on it with me, I have some great ideas, but I'm lazy. Oh, somehow I'm John is lazy. No entrepreneur's lazy, I know that. You're being humble. He is, Haseeb, Kevin, thank you so much for joining me today talking about what you guys are doing at Rafe with EKS, the power why you shouldn't hate K8s. We appreciate your insights and your time. Thank you very much for having us. Yeah, our pleasure. Thank you, you appreciate it. For John Furrier, I'm Lisa Martin. You're watching theCUBE live from New York City at the AWS NYC Summit. John and I will be right back with our next guest, so stick around.