 Welcome back to SuperCloud 22, our inaugural event. It's a pilot event here in theCUBE studios. We're live and streaming virtually until we do it in person maybe next year. I'm John Furrier, host of theCUBE with Dave Vellante. Two great guests, distinguished engineers, managers, CTOs, investors, Marina Tessels, the CTO of Intuit, Insik Ray founder of Vertex Ventures, both have a lot of DNA founder, a loud cloud here with Marc Andreessen, Ben Horowitz, a variety of other great ventures you've done and now you're an investor. Yep, you've been a seasoned CTO, VP of engineering, VMware, Docker, Intuit now, thanks for joining us. Absolutely. So SuperCloud is a thing and apparently it's got a lot of momentum and you guys got SaaS over there at Intuit, Insik, you're investing. And we were challenged on SuperCloud. Our initial thesis was you build on the clouds, get all that leverage like snowflake, you get a good differentiation and then you compete and then move to other clouds. Now it's becoming a thing where I can do this every enterprise, could possibly do it. So I want to get your guys' thoughts on what you think of SuperCloud concept and where are the holes in it, what needs to be defined. Insik, we'll start with you. You've done a lot of cloud things in your day. What do you think? Yeah, I mean it's the whole cloud journey started with a desire to consolidate and desire to actually provide uniformity and standards driven ways of doing things. And I think Amazon was a leader there. They helped kind of teach everybody else. When I was in Cloud Cloud, we were trying to do it with proprietary stacks, just wouldn't work. But once everyone standardized upon Unix and the chipsets no longer became as relevant, they did a lot of good things there. But what's happened since then is now you've got competing standards at the API layer, at the interface layer, no longer at the chipset layer, no longer at the operating system layer. So the evolution of the battles are still there. When you talk about multi-cloud and super-cloud though, like one of the big things you have to keep in mind is latency is not free, latency is very expensive. And it's getting even more expensive now with multi-cloud. So you have to really understand where the separations of boundaries are between your data, your compute. And the network is just there as a facilitator to help binding compute and data, right? And I think there's a lot of bets being made across different vendors like Cloudflare, Akama, as well as Amazon, Google, Microsoft, in terms of how they think we should take computing either to the edge from the core or back and forth. This is structural change. I mean, this is structural. It's desired by incumbents, but it's not something that I'm seeing from the consumption. I'd love to hear from Ariane's perspective from a consumption point of view, like how much edge computing really matters, right? Ariane? So I think there's kind of a story of like two, like it's kind of you can cut it from both edges, not pun intended. On one end, it is really simplifying to actually go into like a single cloud and standardize on it and just have everything there. But I think what over time companies find is that they end up in multiple clouds, whether like through acquisitions or through like needing to use a service in another cloud. So you do find yourself in a situation where you have multi-cloud and you have to kind of work through it and understand how to make it all like work. And latency is an issue, but also for many, many workloads, you can work around it and you can make it work where you have workloads that actually span multiple vendors and clouds. Again, having said that, I would say the world is such that it's still a simplifying assumption when if you go to a single cloud, it's much easier to just go and bet on that. Easier in terms of everything's integrated. IaaS works with SaaS, they solve a lot of problems. Correct, and you can do like for your developers, you can actually provide an environment that's super homogeneous, simple. You can use services easily up and down the stack. And we actually made that deliberate decision when we started migrating to the cloud. At the beginning it was like, oh, let's do like hybrid, we'll make it so it work anywhere. It was so complicated that it was not worth it. When did you give up? What was the moment? Was there a flash point where you said, this is terrible, this is dead? Yeah, when we started to try to make it interoperable and you just see what it requires to do that and the complexity of the architecture that it just became not worth it for the gains you have. So speaking obviously as a SaaS provider, so it just doesn't, it didn't make business case sense for you guys to do that. So it was super cloud then, an infrastructure thing. We just heard from Benoit de Javille that they're going beyond instantiating their data cloud. They're actually running their own little snow grid, they called it. And then when I asked him, well, what about latency? He said, well, we copied data over. So, okay, that's what you have to do. But that's a singular experience with the same governance, with the same security. Just wasn't worth it for you guys is what I'm hearing. Correct, but again, like for some workloads or for some services that we want to use, we are going to go there and we are going to then figure out what is the work around the latency issue. Whether it's like copy or redundancy. Well, the question I have, Dave, on Snowflake is, maybe one question for you and in the panel, is Snowflake a tan expansion opportunity? Or is it a technical reason to go to other clouds? I think they wanted to leverage the hyperscale infrastructure globally. And they said that they're out there. It's a free gift. We're going to go take it. I think it started with, we're on AWS, and then we're on Azure, and then we're on Google. And then they said, why don't we just connect all these and make it a singular experience? And yeah, I guess it's a tan expansion as a differentiator and it adds value, right? If I can share data across that global network. We have customers on Azure now, right? Yeah, of course. You guys don't need to go. What do you think about that? Well, I think Snowflake's in a good position because they work mostly with analytical workloads. And you have capacity that's always going to increase. Like no one subtracts your analytical workload like ever. So there's just compounded growth is like 50% or 80% for many enterprises. Despite their best intentions not to collect more data, they just can't stop doing it. So it's different than if you're like an Oracle or a transactional database where you don't have those kind of infinite growth paths. So Snowflake's going to continue to expand footprint and their customers, they don't mind as long as they can figure out the lowest cost denominator for that. Yeah, so it makes sense to be in all the clouds for them. For them for sure. But Oracle just announced with Microsoft what I would call a super cloud, a cross cloud database service running on OCI and Azure with very low latency and a database that looks like a singular experience. Yeah. With a pass layer. That lost me after OCI. That's. That's. Okay, you know, but that's the BS answer for all UVCs that nobody develops on Oracle. Well, it's a $240 billion market cap company. Show me. You all want to be Oracle. We're going to talk about SRDF and EMC next. You all want Oracle. So there we go. You throw that in too. You all want Oracle to buy your companies, your funding, you know, because we all want to be like Oracle with that kind of cash flow. But anyway. Here's one thing that I'm noticing that is going to be really practical, I think for companies that do run SaaS is because like, you know, you have all these solutions, whether it's like analytics or like monitoring or logging or whatever. And each one of them is very data hungry. And all of them have like SaaS solutions that end up copying the data, moving data to their cloud. And then they might charge you by the size of your data. It does become kind of overwhelming for companies to use that many tools and basically maybe have that data kind of charge for it in multiple places because you use it for different purposes. Or just in general, if you have a lot of data, you know, that that is becoming an issue. So that's something that I've noticed in our own kind of, you know, a world, but it's just something that I think companies need to think about how they solve because eventually a lot of companies will say I cannot have all these solutions or there's no way I'm going to be willing to have so many copies of the data and actually pay for that so many times. Just something to think about. But one of the criticisms of the super cloud concept is that it's just SaaS. If I'm running workload on-prem and I've got, you know, connection to the cloud, which you probably do, that's SaaS, what's the big deal? And that's not anything new or different. So I'd love to get your thoughts on that, but Goldman Sachs, for instance, just announced the service Last Reinvent with AWS connecting their tools, their data and their software from on-prem to AWS. They're offering it as a service. I'm like, hmm, kind of looking like super cloud, but maybe it's just SaaS. It could be. And like what I'm talking about is not so much like, you know, like whether you want to connect your data, but the idea is like a lot of the providers of different services, like in the pads and like higher layer, they're actually copied the data or they need the data in their cloud or their solution. And it just becomes complicated and expensive is kind of like my point. So yes, connecting it like for you to have the data in one place and then be able to connect to it. I think that is a valid, if that's kind of what you think about as a super cloud, that is a valid need. I think that companies will have. Where developers actually want access to tools that might exist. So the key is developers, right? Yeah, absolutely. Developers decide all decisions, not database on administrators, not, you know. 100%. Security engineers, not as admins. So what's really interesting is where are the developers going next? If you look at the current winners in the current ecosystem, companies like MongoDB, and they capture the minds of the JavaScript, you know, Node.js developers. Absolutely. And I started Couchbase and I can tell you like, the difference was that capture motion was so important. So developers are basically used to this game-like experience now where they want to see tools that are free. Whether it's open source or not, they actually don't care. And they want it SAS. They want it SAS, deliver it on demand, right? And pay as you go. And so there's a lot of these different frameworks coming out, next generation, no code, low code, whether it's Java, JavaScript, Rust, you know, whatever, you know, Golang. And there's a lot of people fighting religious wars about how to develop the next kind of modern pattern, design pattern. And that's where a lot of the excitement is how we look at like investment opportunities like. Where are the big bets? Who are frustrated developers? Why are they frustrated? What's wrong with their current environment? You know, do they really enjoy using Kubernetes or trying to use Kubernetes? Right, like developers have a very different view than operators. Well, you mentioned Couchbase. I mean, I look at Couchbase, what you're doing with Capellus as a form of super cloud. I mean, I think that's an excellent, they're bringing that out to the edge. We're going to hear later on from someone from Couchbase that is going to talk about that now. It's kind of a lightweight, you know, sort of it's going to be a synchronization, but it's the beginning. The cool new venture deal that I'm not in, but it was like DuckDB, I'm like, what's DuckDB? Like, well, it's an in-memory database that has like this like remote store thing. I'm like, okay, that sounds interesting. Like, let's call Mike Olson, because that sounds like Sleepy Cat, we don't redistribute the world. But like, it's like, there's a lot of people refactoring design patterns that we're all grew up with since the PubSub days of, you know, take a rendezvous, right? Yeah, that's the refactoring. I think that's the big pattern. So I have to ask you guys, what are you guys investing in? We've got a couple of minutes left to chat about that. What are you investing at into it from a CTO engineering perspective? And what are you investing in that feels super cloud like to you? Well, the thing that like I'm focused on is to make sure that we have absolutely the best in the world development environment for our engineers, where it's modern, it's easy to use, and it incorporates as many things as we can into that environment. So the engineers don't have to think about it. Like one big example would be security and how we incorporated that into development environments. So again, the engineers don't have to bother with trying to think through how they secure their workloads and every step of the way. There are other things that we incorporated, whether it's like rollbacks or monitoring or, you know, like business of other things. But I think that's really an investment that has panned off for us. We actually started investing in development environments over years ago. We started to measure our development velocity and it actually went up by six X just by investing in things like that. So user experience, developer experience and productivity, pretty much, right? Yeah, absolutely. Yeah, that's like a big investment area for us, that, you know, cloud-related. Sounds like a super-cloud-like factor. So many of you are on AWS. We are mostly on AWS, yes. And, Sik, what are you investing in from a VC money doling out standpoint that feels super-cloud-like? So very similar to what we just touched on a lot of developer tool experiences. We have a company that we've invested in cut-offs level that does service catalogs. It's helping, you know, understand where your services live and how they could be accessed. And, you know, enterprise kind of elities that come with that. And then we have a company called Lumigo that helps you do serverless debugging, container debugging, because it turns out debugging distributed, you know, applications is a real problem right now. Just, you can only do so much by log tracing, right? We have a company, haven't announced yet, that's in the WebAssembly space. So we're looking at modernizing the next generation pass stack and throwing everything out the window, including Java and all of the, you know, current pre-built components, because it turns out 90% of enterprise workloads are actually not used. They're just positive code you've compiled with, they're sitting there as vulnerabilities that no one's actually accessing, but you still have to compile with all of it. So we have a lot of bloatware happening in enterprise. So we're thinking about how do you skinny that up with the next generation pass? That's enterprise capable, with security context and frameworks. Super pass. Well, yeah, super pass. That's a kind of good way to put it. Well, is it a consistent developer experience across clouds? It is, and WebAssembly is a very raw standard. If you can call it that, I mean, but it's supported by every modern browser, every major platform vendor, Cloudflare and Adobe and others, and are using it for their own uses. And it's not just about your edge browser compute. It's really, you can take the same framework and compile it down to server side as well as client side. Just like JavaScript was a client side tool before it became Node, right? So we're looking at that as a very interesting opportunity. It's very nascent. Yeah. Great patterns. Yeah. Well, thanks so much for speaking in time out of your busy day. Marianna, thanks for your commentary and appreciate your coming on theCUBE's first and all girl super cloud event pilot. Thanks for sharing. Thanks for having us. Thanks for having us. Okay. More coverage here, SuperCloud 2022. I'm Trevor, Dave Vellante. Stay with us. We've got our Clouderati panel coming up next.