 Welcome to this CUBE Conversation. I'm Dave Nicholson and this is part of the AWS startup showcase season two. I'm very happy to have with me, Webb Brown, CEO of CUBE Cost. Webb, welcome to the program. How are you? I'm doing great. It's great to be here, Dave. Thank you so much for having me. Really excited for the discussion. Good to see you. I guess we saw each other last down in Los Angeles for CUBECon, right? Exactly right. Still feeling the energy from that event. Hoping we can be back together in person, not too long from now. Yeah, well, I'll second that. Well, let's get straight to it. Tell us about CUBE Cost. What do you guys do? And I think just central to that question is, what gives you guys the right to exist? What problem are you solving? Yeah, I love the question. So first and foremost, CUBE Cost, we provide cost monitoring and cost management solutions for teams running Kubernetes or cloud native workloads. Everything we do is built on open source. Our founding team was working on infrastructure monitoring solutions at Google before this. And what we saw was, as we had several teammates showing the Kubernetes effort very early days at Google, we saw teams really struggling even just to monitor and understand Kubernetes costs, right? There's lots of complexity with the Kubernetes scheduler and being able to answer the question of, what is the cost of an application or what is the cost of a team department, et cetera, and the workloads that they're deploying was really hard for most teams. If you look at CNCS study from late last year, still today about two thirds of teams can't answer where they are spending money. And what we saw when digging in there is that when you can't answer that question, it's really hard to be efficient. And by be efficient, we mean get the right balance between cost and performance and reliability. So we help teams in these areas and more. We now have thousands of teams using our product. We feel we're just getting started on our mission as well. So when people think of Kube cost, they naturally associate that with Kubernetes and they think, well, Kubernetes is open source. Wait, isn't that free? So what costs are you tracking exactly? Yeah, great question. We would track costs in any environment where you can run Kubernetes. So if that's on-prem, you can bring a custom pricing sheet to monitor, say the cost of your underlying CPU cores, GPUs, memory, et cetera. If you're running in a cloud environment, we have integrations with Azure GCP and AWS where we would be able to reflect all the complexity of whatever deployment you have, whether you're using Spot in multiple regions where you have complex enterprise discounts, RIs, savings plans, you name it, we'd be reflecting it. So it's really about not just generic prices, it's about getting the right price for your organization. So the infrastructure that goes into this calculation can be on-premises or off-premises in the form of cloud. I heard that right? Yeah, that's exactly right. So all of those environments, we'd give you visibility into all the resources that your Kubernetes clusters are consuming. Again, that's nodes, load balancers, every resource that it's directly touching, also have the ability for you to pull in external costs. So if you have Kubernetes tenants that are using S3 or Cloud SQL or another external cloud service, we would make that connection for you. And then lastly, if you have shared costs, sometimes even like the cost of a DevOps team, we'd give you the ability to kind of allocate that back to your core infrastructure, which may be used for showback or even chargeback across your organization. So who are the folks in an organization that are tapping into this? Are these, are our developers being encouraged to be cognizant of these costs throughout the process or is this just sort of a CFO on down visibility tool? Yeah, it's a great question. And what we see is a major transformation here, where kind of shift left from a cost perspective where more and more engineering teams are interested in just being aware or having transparency. So they can build a culture of accountability with costs, right? With the amazing ability to rapidly push to production and iterate with microservices and Kubernetes, it's hard to have this kind of, just wait for say the finance team to review this at the end of the month or the end of the quarter. We see this increasingly being viewed in real time by infrastructure teams, by engineering teams. Now, finance is still a very important stakeholder and absolutely has a very important like seat at the table in these conversations, but increasingly these are, again, real time or near real time engineering decisions that are really moving the needle on costs and cost efficiency over time and performance as well. Now, can you use this to model what costs might be? Or is this, you mentioned monitoring in real time, is this only for pulling information as it exists? Or could you use some of the aspects of your tool set to make a decision whether something makes more sense to run on your existing infrastructure on-premises versus working in a cloud? Is that something that it's designed for or not? Great question. So we do have the ability to predict cost going forward based on everything we've learned about your environment whether you're a multi-cloud, hybrid cloud, et cetera. So some really interesting functionality there and a lot more coming later this year because we do see more and more teams wanting to model the state of the future, right? As you deploy really complex technologies like say the cluster autoscale or HPA in different environments it can be really challenging to do an apples to apples comparison. And we help teams do exactly that. And again, gonna have a lot more interesting announcements here later this year. So later this year meaning not in the next few minutes while we're together? Nothing new to announce on that front today but I would say expect later this quarter for us to have more to share. Okay, sounds good. Now, you touched on this a little bit but I wanna hone in on why this is particularly relevant now and moving into the future. We've always tracking costs has always been important even before the dawn of cloud but why is it increasingly important? And there are alternatives for cost tracking legacy alternatives that are out there. So talk about why it's particularly relevant now and tell us what your superpower is. You know, what's the secret sauce is something you can't share superpower you can talk about it. Absolutely. And de-ailable so. Yes. So why relevant? What's your superpower? Yeah, great questions. So first of all just to touch on what's fundamentally changing to make yeah, a company like ours impactful or relevant there's really three things here. First and foremost is the new abstractions or complexities that come with Kubernetes, right? Super powerful, but from a cost standpoint make it considerably harder to accurately track costs. And the big transformation here is with Kubernetes you can at any given moment have 50 applications running on a single node or a single VM. You can fast forward five minutes and there could be 50 entirely new applications, right? So just assigning that VM or tagging that VM back to an application or team or department really is not relevant in those places. So just the new complexity related to cost makes this problem harder for teams. Second is what we touch on is just again the power of Kubernetes is the ability to allow distributed engineering teams to work on many microservices concurrently. So you're no longer in a lot of ways managing this problem where they centralize kind of single point of decision making. Oftentimes these decisions are distributed across not only your infrastructure team, but your engineering team. So just the way these decisions and innovation is happening is changing how you manage these. And lastly, it's just scale, right? The cloud and Kubernetes continue to be incredibly successful where as Goop cost now managing billions of dollars, as these numbers get bigger and bigger just becomes more of a business focus and business critical issue. So those are the three kind of underlying themes that are changing. When I talk about what we do that makes this special it's really this like foundational layer of visibility that we built. And what we can do is in real time with a very high degree of accuracy at the largest Kubernetes clusters in the world give you visibility at any dimension. And so from there, you can do things like have real time monitoring. You can have real time insights. You can allow automation to make decisions on these inputs or data feeds. You can set alerts, you can set recurring reports. All of these things are made possible because of the I would say really hard work that we've done to again give this real time visibility with a high degree of accuracy at crazy scale. So if we were to play a little make-believe for a moment pretend like I'm a skeptical sitting on the fence not sure if I wanna go down this path kind of person. And I say, you know what Webb I think I have a really good handle on all of my costs so far. What would you hit me with as an example of something that people really didn't expect until they were running Kube cost and they actually had that visibility? What are some of the things that people are surprised by? Yeah, great question. There'd be a number. Number one I'd have one data point I wanna get from you which is for your organization or for all of your clusters. What is your cost efficiency? Can you answer that with a high degree of accuracy and by cost efficiency? And the answer is no. So tell me how to sign up for Kube cost. Yeah. And so the answer you continue. The answer there is you can go get our community version you can be up and running in minutes. You don't have to share any data like it is simply a helm install. But cost efficiency is this notion of every dollar that you are spending on provision resources. What percentage of those dollars are you actually utilizing? And we now have thousands of teams using our product and we've worked with hundreds of them really closely. This is, that's not the entire market but in our large sample sizes we regularly see teams start in the low 20% cost efficiency. Meaning that approximately 80% is quote waste. Time and time again, we see teams just be shocked by this number. And again, most of it is not because they were measuring it inaccurately or anything like that. Most teams again, today still just don't have that visibility until they start working on this. So is that sort of the, look, in my house household certain members seem to only believe that there is one position for a light switch and that would be the on position. Is this a bit of a parallel where folks are spinning up resources and then just out of sight, out of mind maybe not spinning them down when not needed? Yeah, that's definitely one class of the challenges. I would say, so today, if you look at our product we have 14 different insights across like different dimensions of your infrastructure. One, or I would say several of those relate to exactly what you just described, which is you spin up a VM, you spin up a load balancer, you spin up an external IP address, you're using it, you're not paying for it. Another class is this notion of, again, I don't have an understanding of what my resources cost. I also don't have a great sense for how much my microservice or application will need. So I'm just going to turn on all the light switches or I'm going to drastically over provision. Again, I don't know the cost. So I'm just going to kind of set it and forget it. And if my application is performing, then we're doing well here. Again, with this visibility, you can get much more specific, much more accurate, much more actionable with making that trade-off. Again, down to the individual pod workload, deployment, et cetera. So we've touched on this a bit peripherally, but give me an example. You run into someone who happens to be a happy user of KubeCast. What's the dream story that you love to hear from them about what life was before KubeCast and what life was like after KubeCast? Yeah, there's a lot of different dimensions there. One is working with an infrastructure team that used to get asked these questions a lot about, why does this cost so much or why are we spending this in Kubernetes or why are expenses growing the rate that they are? Like when this works, engineering teams or infrastructure teams aren't getting asked those questions, right? The tool KubeCast itself is getting asked that and answering that. So I think one is infrastructure teams not fielding those types of questions as much. Secondly is just more and more teams rolling this out throughout their organization and ultimately just building a culture of awareness like ownership accountability. And then we just increasingly are seeing teams find this right balance between cost and performance. Again, so in certain cases improving performance when there are resource bottlenecks in places and other places reducing cost by 10 plus million dollars. Ultimately at the end of the day, we like to see just teams being more comfortable running their workloads in Kubernetes, right? That is the ultimate sign of success. It's just an organization feels comfortable with how they're deploying, how they're managing, how they're spending in Kubernetes. Again, whether that be on-prem or transitioning from on-prem to a cloud in multiple clouds, et cetera. So we're talking to you today as part of the second season of the AWS startup showcase. What's the relationship there with AWS? So it is the largest platform for Kube cost being run today. So I believe at this point, at least a thousand different organizations running our product on AWS hosted clusters, whether they're EKS or self-managed, but a growing number of those on EKS. And we've just absolutely loved working with the team across, I think at this point, six or seven different groups from marketplace to their containers team, obviously EKS and others. And just very much see them continuing to push the boundaries on what's possible from a scale and ease of use and just breadth of offerings in this market. Well, we really look forward to having you back and hearing about some of these announcements, things that are coming down the line. So we'll definitely have to touch base in the future. But just one final more general question for you. Where do you see Kubernetes in general going in 2022? Is it sort of linear growth? Is there an inflection point that we see? A good percentage of software that's running enterprises right now is already in that open source category. But what are your thoughts on Kubernetes in 2022? Yeah, I think, the one word is everywhere is where I see Kubernetes in 2022, like very deep in the like large and really complex enterprises, right? So I think you'll see just major bets there. I think you'll continue to see more engineers adopted. And I think you'll also continue to see more and more flavors of it, right? So some teams find that running Kubernetes in a more serverless fashion is right for them. Others find that having full control at every part of the stack, including running their own auto scaler, for example, is really powerful. So I think just you'll see more and more options. And again, I think teams increasingly adopting the right abstraction level on top of Kubernetes that works for their workloads and their organizations. Sounds good. We'll come back in 2023 and we'll check and see how that all panned out. Well, it's been great talking to you today as part of the startup showcase, really appreciate it. Great to see you again. It's right about the time where I can still tell you happy new year because we're still in January here. Hope you have a great 2022. With that, for me, Dave Nicholson, part of theCUBE, part of AWS startup showcase season two, I'd like to thank everyone for joining and stay with us for the best in hybrid tech coverage.