 Welcome back to theCUBE's coverage of KubeCon CloudNativeCon 21, live from Los Angeles, Lisa Martin with Dave Nicholson. And we've got a CUBE alum back with us, Webb Brown is back, the co-founder and CEO of KubeCost. Webb, welcome back. Thank you so much. It is great to be back. It's been right at two years. A lot's happening in our community and ecosystem as well as with our open source project and company. So awesome to be back. So give the audience an overview in case they're not familiar with KubeCost and then talk to us about this explosive growth that you've seen since we last saw you in person. Yeah, absolutely. So KubeCost provides cost management solutions, purpose built for teams running Kubernetes and CloudNative. So everything we do is built on open source. All of our products can be installed in minutes. We give teams visibility in a span then help them optimize it and govern it over time. So it's been a busy two years since we last talked. We have grown the team about 5X. So right around 20 people today. We now have thousands of mostly medium and large size enterprises using the product. That's north of a 10X growth since we launched just before KubeCon San Diego. Now managing billions of dollars of spin and feel like we're just getting started. So it's an incredibly exciting time for us as a company and also just great to be back in person with our friends in the community. This community is such a strong community and it's great to see people back here, I agree. Absolutely, absolutely. So KubeCost, obviously you talk about cost optimization but you're an insight engine in the sense that if you're looking at cost, you have to measure that against what you're getting for that cost. So what are some of the insights that your platform or that your tool set provides? Yeah, absolutely. So we think about our product as first and foremost like visibility and monitoring and then insights and optimization and then governance. If you talk to most teams today, they're still kind of getting that visibility but once you do, it quickly leads into how do we optimize? And there we're going to give you insights at every part of the staff, right? So like at the infrastructure layer, thinking about things like spot and RIs and savings plans, et cetera. At the Kubernetes orchestration layer, thinking about things like auto-scaling and setting requests and limits, et cetera all the way up to like the application layer with all of that being purpose-built for cloud-native Kubernetes. So the way we work is you deploy our product in your environment, anywhere you're running Kubernetes, one dot 11 or above will run and we're going to start dynamically generating these insights in minutes. And they're real-time and again, they scale to the largest Kubernetes clusters in the world. You said you have a thousand or so customers in the medium to large enterprise. These are large organizations, probably brand names I imagine we're familiar with that are leaning on Kubecast to help get that visibility that before they did not have the ability to get. Absolutely, absolutely. So definitely our users of our thousands of users skews heavily towards medium and large-sized enterprise. Working with some amazing companies like Adobe who just have such high-scale and like complex and sophisticated infrastructure. So I think this is very natural in what we expect which is like as you start spending more resources, missing visibility, having unoptimized infrastructure starts to be more costly. Absolutely. And we typically see is once that gets into like the multiple headcount, right? It starts to make sense to spend some time optimizing and monitoring and putting a learning in place so you can manage it more effectively as time goes on. Do you have any metrics or any X factor ranges of the cost that you've actually saved customers? Yeah, I mean, we've saved multiple customers in them like many of millions of dollars at this point. So we're talking big, big, big things. So yeah, we're now managing more than $2 billion of spend. So like some really big savings on a per customer base but it's really common where we're saving north of 30%, sometimes up to 70% on your Kubernetes and related spend. And so we're giving you insights into your Kubernetes cluster and again the full stack there but also giving you visibility insights into external things like external disk or cloud storage buckets or cloud SQL that sort of stuff, external cloud services. Taking us blinders off. Exactly, and giving you that unified real time picture again that accurately reflects everything that's going on in your body. So when these insights are produced or revealed are the responses automated or are they then manually applied? Yeah, that's a great question. We support both and we support both in different ways. By default, when you deploy KubeCost and again today it's Helm install it can be running in your cluster in minutes or less. It's deployed in read-only mode. And by the way, you don't share any data externally it's all in your local environment. So we start generating these insights right when you install and you're involved. And let me ask about, I'm sorry to interrupt it but when you say you're generating an insight are you just giving an answer and guidance or are you providing the reader background on what leads to that insight? Is that a philosophical question of do you need to provide the user rationale for the insight? Yeah, absolutely. And I think we're doing this today and we'll do more but one example is, if you just look at this notion of setting requests and limits for your applications and Kubernetes in simple forms, if you set a request too high you're potentially wasting money because the Kubernetes scheduler is prisoning that resource for you. If you set it too low you're at risk of being CPU throttled. So communicating that symbiotic relationship and the risk on either side really helps team understand why I need to strike this balance. It's not just costs it's performance and reliability as well. So absolutely giving that background and again out of the box we're read-only but we also have automation in our product with our cluster controller so you can dynamically do things like right size your infrastructure or move workloads to spot, et cetera. But we also have integrations with a bunch of tooling in this ecosystem. So like Prometheus native, alert manager native just launched an integration with Spinnaker and Armory where you can dynamically at the time of deployment right size and have insights. So you can expect to see more from us there but we very much think about automation as twofold. One, building trust and coup costs and our insights and adopting them over time. But then two is meeting you where you are with your existing tooling. Whether it's your CI-CD pipeline, observability or existing kind of workflow automation system. Meeting customers where they are is critical these days. Absolutely I think especially in this market where we have the potential to have so much interoperability and all these things working in harmony. And also there's a lot of booths back here. So we have complex tech stacks and in certain cases we feel like when we bring you to our UI or APIs or automation or CLIs, we can do things more effective. But oftentimes when we bring that data to you we can be more effective. Again, bringing your data to Chronosphere or Prometheus or Grafana, all of the tooling that you're already using on a daily regular basis. Bringing that data to the tool is just another example of the value in data that if organizations can actually harness that value and unlock it. There's so much potential there for them to be more competitive for them to be able to develop products and services faster. Absolutely. Yeah, I think you're just seeing the coming of age with cost metrics into that equation. We now live in a world with Kubernetes as this amazing innovation platform where as an engineer I can go spin up some pretty costly resources really fast. And that's a great thing for innovation, right? But it also kind of pushes some of the accountability or awareness down to the individual IC who needs to be aware what things generally cost at a minimum in like a directional way. So they can make informed decisions again when they think about this cost performance reliability trade off. Where are your customer conversations? Are your target users DevOps folks? I'm just wondering where finance might be in this whole game. Yeah, it's a great question. Given the fact that we are kind of open source first and started with open source, we 95% of the time start working with an infrastructure engineering team or DevOps team. They've already installed our product. They're already familiar with what we're doing. But then increasingly and increasingly fast, finance is being brought into the equation and management is being brought into the equation. And I think it's a function of what we were talking about where 70% of teams grew their Kubernetes spend over the last year, 20% of them more than doubled. So these are starting to be real expense items where finance is increasingly aware of what's going on. So yeah, they're coming into the picture but it's simply starting with and working with the infrastructure team that's actually kind of putting some of these insights into action or hooking us into their pipelines. When you think of developers going out and grabbing resources and you think of an insight tool that looks at controlling cost, that could seem like an inhibitor but really if you're talking about how to efficiently use whatever resources you have to have access to in terms of dollars, you could sell this to the developers on that basis. It's like, look, you have these 10 things that you want to be able to do. If you don't optimize using a tool like this, you're only going to be able to do four of them. Without a doubt, yeah. And us as our founding team, all engineers, we were the ones getting those questions of, how have we already spent our budget on just this project? We have these three others we want to do, right? Or why are costs going up as quickly as they are? What are we spending on this application? Instead of that kind of being a manual lift, like let me go do a bunch of analysis or come back with answers, it's tools to where not only can management answer those questions themselves, but the engineering teams can make informed opportunity costs and optimizations decisions as well, whether it's tooling and automation doing it for them or them applying things directly to themselves. So a lot of growth you talked about, the growth in employees, the growth in revenue, what lies ahead for KubeCast? What are some of the things that are coming on the horizon that you're really excited about? Yeah, we very much feel like we're just getting started. Just like we feel this ecosystem and community is, right? Like there's been tons of progress all around, but like, wow, it's still early days. So we did raise $5.5 million from first round who's an amazing group to work with at the end of last year. So by growing the engineering team, we're able to do a lot more. We got a bunch of really big things coming across all parts of our product. You can think about one thing we're really excited that's in limited availability right now is our first hosted solution, it's our first SaaS solution. And this is critically important to us in that we want to give teams the option to, if you want to own and control your data and never egress anything outside of your cluster, you can do that with our deployed product. You can do that with our open source. You can truly lock down namespace egress and never send a byte out. Or if you'd like the convenience of us to manage it for you and be kind of stewards of your data, we're going to offer a great offering there too. So that's in limited availability today. We're going to have a lot more announcements coming there. But we see those being at feature parity between our enterprise offerings and our hosted solution. And just a lot more coming with visibility, some more GPU insights metrics coming quickly, a lot more with automation coming, and then more integrations for governance. Again, kind of talked about Spinnaker and things like that. A lot more really interesting ones coming. So five and a half million raised in the last round of funding. Where are you going to be applying that? What are some of the growth in the engines that you want to tune with that money? Yeah, so first and foremost, it was really growing the engineering team. So we've forexed the engineering team in the last year and just have an amazing group of engineers. We want to continue to do that. We're kind of super early on the marketing sale side. We're going to start thinking about that more and more. Our approach first off was like, we want to solve a really valuable problem and doing it in a way that is super compelling. And we think that when you do that, good things happen. I think that's some of our Google background, which is like, you build a great search engine and good things generally happen. So we're just super focused on, again, working with great users, building great products that meet them where they are and solve problems that are really important to them. Well, congratulations on all the trajectory of success since we last saw you in person. Great to have you back on the show. Looking forward to it. So folks can go to kubcos.com to learn more and see if someone has announcements coming down the pike. Absolutely. Don't make it two years before you come back. I would love to be back. I hope we're back bigger than ever next year. But it has been such a pleasure last time and this time. Thank you so much for having me. Love being part of the show and the community at large. It's a great community. And we appreciate you sharing all your insights. Thank you so much. All right. For Dave Nicholson, I'm Lisa Martin coming to you live from Los Angeles. This is the cubes coverage of KubeCon and CloudNativeCon 21. We'll be back with our next guest shortly. We'll see you then.