>>Good morning and welcome back to the Cube where we are excited to be broadcasting live all week from Detroit to Michigan at Cuban slash cloud Native con. Depending on who you're asking, Lisa, it's day two things are buzzing. How are you feeling? >>Good, excited. Ready for day two, ready to have more great conversations to see how this community is expanding, how it's evolving, and how it's really supporting it itself. >>Yeah, Yeah. This is a very supportive community. Something we talked a lot about. And speaking of community, we've got some very bold and brave folks over here. We've got this CTO and the head of product from Storm Forge, and they are on a mission to automate Kubernetes. Now automatic and Kubernetes are not words that go in the same sentence very often, so please welcome Patrick and Yasmin. Thank you both for being here. Hello. How you doing? >>Thanks for having us. >>Thanks for having us. >>Talk about what you guys are doing. Cause as you said, Kubernetes auto spelling is anything but auto. >>Yeah. >>The, what are some of the challenges? How do you help >>Eliminate this? Yeah, so the mission at Storm Forge is primarily automatic resource configuration and optimization essentially. So we started as a machine learning company first. And it's kind of an interesting story cuz we're one of those startups that has pivoted a few times. And so we were running our machine learning workloads. Most >>Have, I think, >>Right? Yeah. Yeah. We were, we started out running our machine learning workloads and moving them into Kubernetes. And then we weren't quite sure how to correctly adjust and size our containers. And so our ML team, we've got three PhDs and applied mathematics. They said, Well, hang on, we could write an algorithm for that. And so they did. And then, Oh, I love this. Yeah. And then we said, Well holy cow, that's actually really useful. I wonder if other people would like that. And that's kind of where we got our start. >>You solved your own problem and then you built a business >>Around it. Yeah, exactly. >>That is fantastic. Is, is that driving product development at Storm Forge still? That kind of attitude? >>I mean that kind of attitude definitely drives product development, but we're, you know, balancing that with what the users are, the challenges that they have, especially at large scale. We deal with a lot of large enterprises and for us as a startup, we can relate to the problems that come with Kubernetes when you're trying to scale it. But when you're talking about the scale of some of these larger enterprises, it's just a different mentality. So we're trying to balance that of how we take that input into how we build our product. Talk >>About that, like the, the end user input and how you're taking that in, because of course it's only going to be a, you know, more of a symbiotic relationship when that customer feedback is taken and >>Acted on. Yeah, totally. And for us, because we use machine learning, it's a lot of building confidence with our users. So making sure that they understand how we look at the data, how we come up with the recommendations, and actually deploy those changes in their environment. There's a lot of trust that needs to be built there. So being able to go back to our users and say, Okay, we're presenting you this type of data, give us your feedback and building it alongside them has helped a lot in these >>Relationships. Absolutely. You said the word trust, and that's something that we talk about at every >>Show. I was gonna jump on that too. It's >>Not, Yeah, it's not a buzzword. It's not, It shouldn't be. Yeah. It really should be, I wanna say lived and breathed, but that's probably grammatically incorrect. >>We're not a gram show. It's okay darling. Yeah, thank >>You. It should be truly embodied. >>Yeah. And I, I think it's, it's not even unique to just what we do, but across tech in general, right? Like when I talk about SRE and building SRE teams, one of the things I mentioned is you have to build that trust first. And with machine learning, I think it can be really difficult too for a couple different reasons. Like one, it tends to be a black box if it's actually true machine learning. Totally. Which ours is. But the other piece that we run into. Yeah. And the other piece we run into though is, is what I was an executive at United Health Group before I joined Storm Forge. And I would get companies that would come to me and try to sell me machine learning and I would kind of look at it and say, Well no, that's just a basic decision tree. Or like, that's a super basic whole winter forecast, right? Like that's not actually machine learning. And that's one of the things that we actually find ourselves kind of battling a little bit when we talk about what we do in building that trust. >>Talk a little bit about the latest release as you guys had a very active September. Here we are. And towards the, I think end of October. Yeah. What are some of the, the new things that have come out? New integrations, new partnerships. Give us a scoop on that. >>Yeah, well I guess I'll start and then I'll probably hand it over to you. But like the, the big thing for us is we talked about automating Kubernetes in the very beginning, right? Like Kubernetes has got a vpa it's >>A wild sentence anyway. Yeah, yeah. >>It it >>Has. We're not gonna get over at the whole show. Yeah. >>It as a VPA built in, it has an HPA built in and, and when you look at the data and even when you read the documentation from Google, it explicitly says never the two should meet. Right. Because you'll end up thrashing and they'll fight each other. Well the big release we just announced is with our machine learning, we can now do both. And so we vertically scale your pods to the correct up. Yeah. >>Follow status. I love that. >>Yeah, we can, we can scale your pods to the correct size and still allow you to enable the HPA and we'll make recommendations for your scaling points and your thresholds on the HPA as well so that they can work together to really truly maximize your efficiency that without sacrificing your performance and your reliability of the applications that you're running. That >>Sounds like a massive differentiator for >>Storm launch, which I would say it is. Yeah. I think as far as I know, we're the first in the industry that can do this. Yeah. >>And >>From very singularity vibes too. You know, the machines are learning, teaching themselves and doing it all automatically. Yep. Gets me very >>Excited. >>Yeah, absolutely. And from a customer demand perspective, what's the feedback been? Yeah, it's been a few >>Weeks. Yeah, it's been really great actually. And a lot of why we went down this path was user driven because they're doing horizontal scale and they want to be able to vertically size as they're scaling. So if you put yourself in the shoes of someone that's configuring Kubernetes, you're usually guessing on what you're setting your CPU requests and limits do. But horizontal scale makes sense. You're either adding more things or removing more things. And so once they actually are scaled out as a large environment and they have to rethink, how am I gonna resize this now? It's just not possible. It's so many thousands of settings across all the different environments and you're only thinking about CPU memory, You're not thinking about a lot of things. It's just, but once you scale that out, it's a big challenge. So they came to us and said, Okay, you're doing, cuz we were doing vertical scaling before and now we enable vertical and horizontal. And so they came to us and said, I love what you're doing about right sizing, but we wanna be able to do this while also horizontally scaling. And so the way that our software works is we give you the recommendations for what the setting should be and then allow Kubernetes to continue to add and remove replicas as needed. So it's not like we're going in and making changes to Kubernetes, but we make changes to the configuration settings so that it's the most optimal from a resource perspective. >>Efficiency has been a real big theme of the show. Yeah. And it's clear that that's a focus for you. Everyone here wants to do more faster Of course. And innovation, that's the thing to do that sometimes we need partners. You just announced an integration with Datadog. Tell us about that. Yeah, >>Absolutely. Yeah. So the way our platform works is we need data of course, right? So they're, they're a great partner for us and we use them both as an input and an output. So we pull in metrics from Datadog to provide recommendations and we'll actually display all those within the Datadog portal. Cause we have a lot of users that are like, Look, Datadog's my single pane of glass and I hate using that word, but they get all their insights there. They can see their recommendations and then actually go deploy those. Whether they wanna automatically have the recommendations deployed or go in and actually push a button. >>So give me an example of a customer that is using the, the new release and some of the business outcomes they're achieving. I imagine one of the things that you're enabling is just closing that ES skills gap. But from a business level perspective, how are they gaining like competitive advantages to be able to get products to market faster, for example? >>Yeah, so one of the customers that was actually part of our press release and launch and spoke about us at a webinar, they are a SaaS product and deal with really bursty workloads. And so their cloud costs have been growing 40% year over year. And their platform engineering team is basically enabled to provide the automation for developers and in their environment, but also to reduce those costs. So they want to, it's that trade off of resiliency and cost performance. And so they came to us and said, Look, we know we're over provisioned, but we don't know how to tackle that problem without throwing tons of humans at the problem. And so we worked with them and just on a single app found 60% savings and we're working now to kind of deploy that across their entire production workload. But that allows them to then go back and get more out of the, the budget that they already have and they can kind of reallocate that in other areas, >>Right? So there can be chop line and bottom >>Line impact. Yeah. And I, I think there's some really direct impact to the carbon emissions of an organization as well. That's a good point. When you can reduce your compute consumption by 60%. >>I love this. We haven't talked about this at all during the show. Yeah. And I'm really glad that you brought this up. All of the things that power this use energy. Yeah. >>What is it like seven to 8% of all electricity in the world is consumed by data centers. Like it's crazy. Yeah. Yeah. And so like that's wild. Yeah. Yeah. So being able to make a reduction in impact there too, especially with organizations that are trying to sign green pledges and everything else. >>It's hard. Yeah. ESG initiatives are huge. >>Absolut, >>It's >>A whole lot. A lot of companies have ESG initiatives where they can't even go out and do an RFP with a business, Right. If they don't have an actual active starting, impactful ESG program. Yes. Yeah. >>And the RFPs that we have to fill out, we have to tell them how they'll help. >>Yeah. Yes. It's so, yeah, I mean I was really struck when I looked on your website and I saw 54% average cost reduction for Yeah. For your cloud operations. I hadn't even thought about it from a power perspective. Yeah. I mean, imagine if we cut that to 3% of the world's power grid. That is just, that is very compelling. Speaking of compelling and exciting future things, talk to us about what's next? What's got you pumped for 2023 and and what lies >>Ahead? Oh man. Well that seems like a product conversation for sure. >>Well, we're super excited about extending what we do to other platforms, other metrics. So we optimize a lot right now around CPU and memory, but we can also give people insights into, you know, limiting kills, limiting CPU throttling, so extending the metrics. And when you look at hba and horizontal scale today, most of it is done with cpu, but there are some organizations out there that are scaling on custom metrics. So being able to take in more data to provide more recommendations and kind of extend what we can do from an optimization standpoint. >>That's, yeah, that's cool. And what house you most excited on the show floor? Anything? Anything that you've seen? Any keynotes? >>There's, Well, I haven't had a lot of time to go to the keynotes unfortunately, but it's, >>Well, I'm shock you've been busy or something, right? Much your time here. >>I can't imagine why. But no, there's, it's really interesting to see all the vendors that are popping up around Kubernetes focus specifically with security is always something that's really interesting to me. And automating CICD and how they continue to dive into that automation devs, SEC ops continues to be a big thing for a lot of organizations. Yeah. Yeah. >>I I do, I think it's interesting when we marry, Were you guys here last year? >>I was not here. >>No. So at, at the smaller version of this in Los Angeles. Yeah. I, I was really struck because there was still a conversation of whether or not we were all in on Kubernetes as, as kind of a community and a society this year. And I'm curious if you feel this way too. Everyone feels committed. Yeah. Yeah. I I I feel like there's no question that Kubernetes is the tool that we are gonna be using. >>Yeah. I I think so. And I think a lot of that is actually being unlocked by some of these vendors that are being partners and helping people get the most outta Kubernetes, you know, especially at the larger enterprise organizations. Like they want to do it, but the skills gap is a very real problem. Right. And so figuring out, like Jasmine talked about figuring out how do we, you know, optimize or set up the correct settings without throwing thousands of humans at it. Never mind the fact you'll never find a thousand people that wanna do that all day every day. >>I was gonna, It's a fold endeavor for those >>People study, right? Yeah. And, and being able to close some of those gaps, whether it's optimization, security, DevOps, C I C D. As we get more of those partners like I just talked about on the floor, then you see more and more enterprises being more open to leaning into Kubernetes a little bit. >>Yeah. Yeah. We've seen, we've had some great conversations the last day and, and today as well with organizations that are history companies like Ford Motor Companies for >>Example. Yeah. Right. >>Just right behind us. One of their EVs and, and it's, they're becoming technology companies that happen to do cars or home >>Here. I had a nice job with 'em this morning. Yes. With that storyline, honestly. >>Yes. That when we now have such a different lens into these organizations, how they're using technologies, advanced technologies, Kubernetes, et cetera, to really become data companies. Yeah. Because they have to be, well the consumers on the other end expect a Home Depot or a Ford or whomever or your bank Yeah. To know who you are. I want the information right here whenever I need it so I can do the transaction I need and I want you to also deliver me information that is relevant to me. Yeah. Because there, there's no patience anymore. Yeah. >>And we partner with a lot of big FinTech companies and it's, it's very much that. It's like how do we continue to optimize? But then as they look at transitioning off of older organizations and capabilities, whether that's, they have a physical data center that's racked to the gills and they can't do anything about that, so they wanna move to cloud or they're just dipping their toe into even private cloud with Kubernetes in their own instances. A lot of it is how do we do this right? Like how do we lean in and, Yeah. >>Yeah. Well I think you said it really well that the debate seems to be over in terms of do we go in on Kubernetes? That that was a theme that I think we felt that yesterday, even on on day one of the keynotes. The community seems to be just craving more. I think that was another thing that we felt yesterday was all of the contributors and the collaborators, people want to be able to help drive this community forward because it's, it's a flywheel of symbiosis for all of the vendors here. The maintainers and, and really businesses in any industry can benefit. >>Yeah. It's super validating. I mean if you just look at the floor, there's like 20 different booths that talk about cost reporting for Kubernetes. So not only have people moved, but now they're dealing with those challenges at scale. And I think for us it's very validating because there's so many vendors that are looking into the reporting of this and showing you the problem that you have. And then where we can help is, okay, now you know, you have a problem, here's how we can fix it for you. >>Yeah. Yeah. That, that sort of dealing with challenges at scale that you set, I think that's also what we're hearing. Yeah. And seeing and feeling on the show floor. >>Yeah, absolutely. >>What can folks see and, and touch and feel in your booth? >>We have some demos there you can play around with the product. We're giving away a Lego set so we've let >>Gotta gets >>Are right now we're gonna have to get some Lego, We do a swag segment at the end of the day every day. Now we've >>Some cool socks. >>Yep. Socks are hot. Let's, let's actually talk about scale internally as our closing question. What's going on at Storm Forge? If someone's watching right now, they're excited. Are you hiring? We are hiring. Yeah. How can they stalk you? What's the >>School? Absolutely. So you can check us out on Storm forge.io. We're certainly hiring across the engineering organization. We're hiring across the UX a product organization. We're dealing, like I said, we've got some really big customers that we're, we're working through with some really fun challenges. And we're looking to continue to build on what we do and do new innovative things like especially cuz like I said, we are a machine learning organization first. And so for me it's like how do I collect all the data that I can and then let's find out what's interesting in there that we can help people with. Whether that's cpu, memory, custom metrics, like as said, preventing kills, driving availability, reliability, What can we do to, to kind of make a little bit more transparent the stuff that's going on underneath the covers in Kubernetes for the decision makers in these organizations. >>Yes. Transparency is a goal of >>Many. >>Yeah, absolutely. Well, and you mentioned fun. If this conversation is any representation, it would be very fun to be working on both of your teams. We, we have a lot of fun Ya. Patrick, thank you so much for joining. Thanks for having us, Lisa, As usual, thanks for being here with me. My pleasure. And thank you to all of you for turning into the Cubes live show from Detroit. My name's Savannah Peterson and we'll be back in a few.