 Hello, welcome back to theCUBE's live coverage here at VMware Explorer 2023. I'm John Furrier, host of theCUBE, and we have a special presentation here instead of our normal CUBE segment. We're going to do a panel roundtable rapid fire discussion with Kindrell, Microsoft, Oracle, and VMware, all bringing it together really on a topic and we're going to unpack it here in real time. And modernizing in a digital first multi-cloud world is the topic. This is the hottest topic as we call it super cloud. As architectures start to change, the world is now looking at new ways to set the table, set the infrastructure and the software stacks to power modern applications. This next gen cloud is here. They have ecosystems, they have large language foundation models. The entire computing paradigm is changing and growing every day thanks to the benefits of cloud and AI, all goodness. Exciting to have Sunil Bagarva, senior vice president of offers at Kindrell, Manty Bhatia, vice president of global systems area of VMware, Andrew Bow, VP of product development, Azure of Microsoft, and Ross Brown, senior VP, Rosemary Cloud ecosystem part of that Oracle. We have a quorum of awesome experts here who are going to debate and talk about the topics and educate you about what's going on in the cloud world. As you got to figure it out, navigating through is going to be challenging with this huge opportunities at the end of the day. There's so much value being created and transferred over from old to new. And we're all here. Gentlemen, thank you for making history on theCUBE. Thank you. Great to be here. Great to be here. So I don't know where to start. We'll start with Kindrell. You guys set the table for the customers. You guys build out solutions. You've been doing it for a long time. VMware has been partner friendly in bringing a unique combination of traversing multiple environments on-premise. You guys love the on-premise environment and edge. We get the cloud for Azure number two and growing fast. I mean, the growth rates are phenomenal. Oracle, the performance and growth of your cloud has really been amazing in the past few years. Well noted. We're all together. At the end of the day, the customers sitting there going, okay, I have a lot of cloud. And how'd we get here? Okay, I use Microsoft for this. I got Amazon for that. I got Oracle for that. I got VMware for that. I just got to make it run. I'm going to have multiple clouds. I think for us where we start is we really have to start with a customer-centric point of view. Whether it's through acquisition and nurture, whether it's different departments within the customer, they've ended up with a multi-cloud scenario and now they have to understand how do they take all this data, all these applications and modernize them. Whether they modernize in place, whether they leverage different SaaS applications and then how are they going to overlay AI on top of that with all this data to feel that generative model and really see the benefits and the wins and the business value from it. And I think that's what's from an Azure standpoint or from VMware or from Kendra or from Oracle. It's a great big world for all of us right now and I think we all have a unique opportunity to bring a lot of value to customers. Absolutely, and John, as you mentioned, this is a typical definition of what we call as the cloud chaos, right? This is exactly what cloud chaos is. I remember last year who went on the keynote and talked about the cloud chaos and that's what our cloud smart approach is about, right? Where are the right workloads and which cloud they belong to and how do we get them there is the cloud smart approach and I think we're all in this world together to bring our customers to that endpoint, basically. I think opportunity is exactly the right word. Multi-cloud presents an opportunity. Now there is an opportunity cost for not leveraging what is available from all the great partners that are available to bring new technologies and all of your investing in new capabilities, whether it is for Generative AI, security, data analytics, new capabilities coming out all the time. The value Kindrel brings to our customers is being able to help understand for their needs and their challenges. What are the right combination of technologies that can deploy the applications on multi-cloud, on the super cloud, ensuring that they get the best of all the investments that our partners make? I think you're bringing up a really key point about being able to solve for multi-cloud is not just a matter of having a great plan and stuff like that, but also really taking time to figure out how they're diverging. And one of the things you're seeing happen is, I'll say it this way, we are still at the beginning of the beginning for the cloud, right? The value of the work that will be built in Generative AI to create new IP and new assets for companies to be able to instrument businesses in an automated way. The value of that work is a magnitude more than the existing work that's either on-prem and in cloud. So the size of what needs to be built often will exceed the capital budget of any company. So as a result, all the cloud providers are having to make choices of where do we invest our dollars? How do we build out infrastructure? How do we build out past layers? And they're differentiating. So that leads to the need for us to work well together like Microsoft and Oracle, 12 of our locations, we set up a connection between our cloud so customers can go between them. Choice among chaos. Yeah, and the thing that's interesting about the multi-cloud and super cloud conversation is you guys just mentioned it and here at the show at VMware Explorer is Raghu said on stage, a little nuanced point, but I'll make, I have a dog whistle for technical terms to get me excited. I heard multi-clouds run time for multi-run time. That sounds like an operating environment. This is an operating environment. That's kind of like that super environment we see happening. So if you believe that distributed computing is here and that's what cloud is, you ask yourself, okay, what's the ecosystem like? Where's AI fit? So these questions start coming up. And right now, Ross kind of brought this up. I want to put out the topic is these are enterprise challenges. That's right. Yeah, cloud 1.0. Okay, startups got in there, Airbnb, great stuff, SaaS has developed. Now security, enterprise, the old enterprise, they make complex problems more complex. In cloud you can't do that. You got to make them simpler and you got to interoperate. So the question on the table is, their enterprise has real needs, security, performance, compliance and AI. You can put it together, you guys build it. What's that look like in your guys mind? How does the enterprise get what they need? Because you can't falter on any one of those categories. I think when you take a look at an enterprise, right? If you look at individual applications, whether you want to look at an application ready, and enterprises have massive estates that are extremely complex. And they need audit ready, self-healing secure environments that are in an AI sense, that their data is theirs. Whether they're using existing model or whether they're building their own model or they're dealing with the cohort of publicly available data. They need to make sure they can keep that data at home. And basically that intelligent at home and save it. But even across all of their applications across multi-cloud, they need to leverage the best of each and every one of us to do what they can do best in areas that give them the best. I hate to use the term bang for their buck, but that's it. And they're going to take a look at what they have. And they have a long standing partnership with VMware. They have great relationship with Kindrel. They have great relationships with Oracle. They have great relationships with Microsoft. And I think it's to all of us to sit down like this and decide, okay, how can they benefit with all of this together? I don't think there ever will be one answer. And that is why the whole approach around ecosystem is very, very important, right? It's like a spider's web, right? You are bringing many things together. You want to make sure that they are all interlocked. You want to make sure that they are all interoperable. And you're all, you know, at the end of the day, what we're trying to do is solve a customer problem, not make it more complex for them, make it simpler for them. It's up to us as providers to build those things for our customers to, you know, to bring that spider's web together. It's hard to have this conversation without pitching a solution. There's so much I want to say. I know that. Feel free to say it. It's open mic night tonight here. Oh, we're gonna all stay friendly. I think Ross's point earlier was really important that all of us are investing differently. Yep. And as a result, interoperability is not automatic. And that's okay. Because the divergent value in all of the technology investments we need can be put together to handle complex problems and make them simple. And to the point you made, John, in terms of making sure that the operating environment, the operating model, operations, any aspect of that continuum remains true to delivering value while leveraging the investments all of you have made. And that's what makes the customer's objectives real. We wing their innovation agenda while capitalizing on all the innovations all of our partners are bringing. Ross, I'll let the ask you a question. I know OCI has always been performance. I've been covering Oracle for a long time. Performance performance, security performance. That's kind of the bread and butter of Oracle. Yeah. When you look at multi-cloud, how do you see that optimizing for you guys as you're in this multi-environment mode and still maintain that performance? Because that's going to come up a lot. Does it perform? And certainly with data in the edge coming. There's a slightly technical answer I won't go into on performance that gets into non-blocking networks and layer two virtualization stuff that it really isn't. As you look at clouds grow, you end up with collisions. You end up with challenges and stuff. We tried to engineer, because we're late. We had the opportunity to look at whether people done and said if you were to start over, what would you do? So we see Azure and other folks making similar investments, but we had the blank slate so we could just start that one, right? That notion of performance is something that's very critical to our exadata platform, our applications and all those things. But it has to be performance at a reasonable price. And I think that's the thing customers are really looking for is when you look across the different clouds, am I getting to your point the most bang for the buck? And part of that equation is, is it easy? And in many cases, that's where multi-cloud becomes really critical because you have skill sets that are in one, you have application of states that are in one, you have data that's sitting in another one. And I look at it this way. All of us share a piece of the wisdom our customers need, right? We all bring to the table a piece of it. And when we collaborate, work well together, we end up delivering a greater outcome for the customer, a wiser result. And I think the pressure from customers for us to do this is higher than it ever has been. I mean, we get conversations from folks of like, why don't you play well with those guys? Well, we're open to it. Let's play well with it because we're a changed oracle in the sense that we realize the opportunity in front of us is massive. And it's not about protecting a database moat or protecting these things. It's about pleasing customers at scale in a way that you can do, right? That's the big point. I remember the 80s and 90s, you know? Multi-vendor. How do you do multi-vendor in the cloud when you have an operating environment, not just buying a solution? They got to work together. API's were great. Now you got LLMs coming in. You got to think about what needs to work together. And this is where I think you have a lot of experience on that, where things get integrated or not. You have silos, like I will make the argument app tiers are isolated, data needs to integrate. So you brought up with APIs, you brought up that's replicating data between systems, being able to have a consistent set of records that you can, immutable records you can count on and those things. That's where the interoperability comes in. Can you move information between these? Because you might be running the GenIA model that my data that's sitting from a transactional system is going into. Okay, well it has to be able to go between, right? And that's where the friction's happening. And that's where the complexity lies as well, right? That's where you have the compliance issues, the legal issues also. Because the data can be anywhere and data is your, basically that's your- It's the asset. It's the asset. It is the new oil. I do, I want to- It's not necessary from an operations perspective. It's something that requires visibility with actionable intelligence. And whether that's the bridge technology from us, the area from VMware, that community is pulling together to bring this capability to our customers. So that products like Bridge can give the insights and the actionable intelligence so the customers can have the confidence that app tier may be vertically isolated, the data is integrated. Policy for compliance is being addressed. There is insight into cost. And we are actually delivering, as we had talked about very early in the conversation, bang for the ball. One last huge point on that, which is the human capital. That companies have invested in Oracle, in Microsoft, in VMware, all of the teams that Kindrel brings to bear with their customers. Companies just can't shift this overnight, right? And I mean, not without massive impact. In today's environment, it's almost impossible to get enough people skilled in scale. You have to leverage the assets you have. That is a very important point that we don't necessarily talk about often. We talk about technology. We don't talk about the skills. There's years and years of skills that have been built up in some of these technologies. And we need to be able to leverage those in this multi-cloud environment. And get the return on the original investments. Yeah, exactly. I want to double click on this. I think this is really worth addressing because it comes into the conversation on multiple clouds working together. The people skills, if AI continues to go on a tear, as we say it's early, and the gift hasn't started to be given yet, well, if it's going to give more, it's going to be great. Cloud, you got VMware, you got Microsoft, you got Oracle, you got people that have specialized on these technologies and the stacks. And now you have, okay, if automation comes in, operations and admin role, they're well understood concepts and could potentially have help with AI. The cloud architect, to me, I think, is coming, I'm hearing out of this event here, cloud architect will be the premium job. It is, it is. Although I just make sure we address one thing real quick, which is cloud skills are the skill. Like knowing how to build an architect to cloud solution takes a while to learn that. If you're an AWS engineer, you could become an Azure engineer like that. It's like I've been trained on Chrysler's my entire life. I understand transmissions and engines. I walk in a Ford shop, I'm going to need to learn the tooling, I'm going to need to learn the specific gauges and stuff, but the principle of the engineering is the same. So that's one of the things we're starting to see is like Kubernetes and containers. So standardized now that unless you're dealing with like a really odd way to run containers like Corey Quinn does and putting them in route 53 or whatever, that's a really odd way to run containers. But if you're running them in a normal way, the skills are managed Kubernetes, the skill managed Kubernetes. And what I start seeing happen is like, you talked about the lack of people. I have this question in my head, which is how many man hours are in terraform right now? How much automation have we already captured that intelligence and how much more will we capture in tooling and scripting and automation? And as we start building machine learning into that automation, I think the human capital side of this, which is by far the largest budget required, will go down. I mean, it will become because we need to reinvest it in the next bow wave. So the interesting thing is skills that are technology based, absolutely what you're saying would happen. But the knowledge about the application of the business, that still remains and it's not easy to harvest digitally. And that's where the role of stitching together multiple cloud architectures, defining policy, defining the operating model for operations, that becomes key and that's not as easy to automate. The data, there's a data, you bring in, it's all about data. You mentioned that. And I remember the old search days, contextual and behavioral data. If you have, and class a little bit different, I want to get you guys reaction on this. And I've been saying this on theCUBE in other interviews. It's different, it's not just behavioral, because behavior describes something. You got contextuals that's still relevant, but the data is now action. Because if automation comes in, actions are happening. So it's context and the behavior is action. Or do you disagree or? I don't necessarily disagree, but when I think of things on application level, all applications do is collect and process data. I mean, if you look at generative AI, I mean, at its base level, it's a natural language interface, a logic interface, chewing through a bunch of data in a way that you directed. But to me, how we aggregate or disaggregate data or how we make it work across different platforms, this is a solution that I think AI can solve for companies and back to the human skilling, the people that can start to really understand how to drive that engine. It's huge value for businesses, it's huge value for companies. Andrew, let me just challenge you a little bit on that part, right? Just because of the fact, because the data is all over the place right now, right? And for the LLMs to work, you really need to aggregate that data for it to really be not hallucinating, right? So chat GPT can hallucinate based on what data is provided to it. And that is still a big issue that has not been solved, or at least not, you know, there's data all over the place, publicly available data, there's private data, and for us to be able to use AI to really make particular use cases work, we need to have that aggregation, and that's a problem that's still existing. It's a growing problem. It's a bigger problem. Well, the unstructured data is the layer of policy on data. Absolutely. You are allowed, what purpose are you allowed to use it for? And security too, right? Oh, absolutely. And the volume of data that's being created, the amount of data created today at the edge, it's just coming at such a volume that we need assistance to process it and make intelligent decisions. And that's, I think, where we're going with a lot of this. And I think this enterprise, back to the enterprise question earlier, real-time data is right now great because we are making action on real-time. The historical data also now is input into decision-making of real-time, and the data is constantly changing, and then the authenticity of the data, the trust of the data comes in. So the question is, on the table is, we see software supply chain, S-bombs and whatnot, what about data supply chain? Oh, absolutely. I mean, that's going to be an issue. I want to point out something. You guys are talking about large language models and sort of interpreting data that's like document data and unstructured and things like that. There are companies that are building right now AI engines that are transformers that are different. In the sense that they're API aware and they're SDK aware. They know how systems work. And I'll just ask this question. If you want to understand a company's strategy, go look at their ERP system and the policies that are in ERP. Their entire strategy is outlined. All their workflows on how you can- It's like DNA. You could take a big print of a company and say, how conservative are they based on the approval policies on expenses, approval policies and POS? So there's an induced information and then there's direct information. That induced information on the systems that we built over time. That's going to be transformative because when you pair that with something that can automatically access APIs and build code, you start getting goal-oriented software systems. Yeah, yeah. I mean, it's almost- When you're going to manufacture data, then the risk of the outcome being on target increases. Yeah, absolutely. Not that you can't manage it, but that just increases the complexity ever more. So are we going to have a creative class in technology? Yeah. Because that's where it's going. It is, it is. If you think what you're happening on automated tasks of John, the creativity, just what you hear on all the themes, what does that look like? What does the creative class look like? Just imagine, if you buy a copy of Photoshop, it'll show you every developer who worked on Photoshop. It was a tremendous piece of engineering to create that and it's exclusively used by artists now. So you say, is there going to be a creative class with technology? I would argue that's the end game for almost all technologies. Yeah, that's awesome. It's been able to- What does it look like? Yeah. What does that look like? What does the creative class look like? Magically, Harry Potter like? No, they're just provisioned the servers. They're more thinking about if I can envision a new business model and I could get coached from an AI on where the gaps in strategy are that I need to tune up. If I could propose a goal and a set of strategies to it, could it map the system to achieve those strategies? That's what I think about creative. Yeah, that's awesome. Or can it find efficiencies? Yeah, yeah. Right? And I mean, I think of things like call centers, right? Where in modern call centers, somebody has a very frustrating experience trying to navigate. Now, AI can take a look at that customer, that customer history and maybe have a somewhat of a logical reason for the call and help direct them to the right person and also prompt the person that they're speaking to with a direction. So it can be a real top-down integration. I think it's very valuable. I mean, the use cases are just, there's so many use cases, right? Predictive maintenance, manufacturing, automotive. You just take it, there's so many use cases where you can use it right now and create so much efficiencies. The most interesting creativity I'm seeing right now is in multi-cloud architectures. Not to bring it home, but multi-cloud architectures create an opportunity for an organization to really capture to your earlier point about what are their policies? What is it? Just define it in the system. And that's why I think cloud architecture is a systems thinking. We hear design thinking. I think we're entering an era of system thinking, creativity, and personalization. Because what you just said, your example was personalization. No one's across that. Map me out my goals, like, go. There's been a massive change in cloud developments and this is true in every cloud. This is not anything in particular. But I think it's been missed and it's, I think, important to put up, which is if you look at a private data center like companies, things are done in layers. You have a five-year network refresh where you do your network architecture, storage refresh. In the cloud, every application can be its own side of the service because it can be unique. So as a result, the creativity that goes into what do I need for this workload can be vastly different than what you're building in the next work in the workload. So that enables a whole new set of developer creativity in the cloud that you're not locked into one architecture in the network. I think that also shifts the way people work, you know, in terms of my job, right? In that example, what we are finding is more customers than ever are adopting a product mindset. And the creativity is not just in the developers with the application or the database, but the creativity is in the entire stack, full process, from the business process down to which clouds, which data, what analytics. My final question to put on the table and we'll go around for closing summary, each of you, if you don't mind. But the final question is, we obviously love general AI, we got general BI coming next probably. All this is going on. You got the dorm room to the dorm, from the board room to the dorm room, activity. T-suite, take that hill, I need AI and everything. Okay, now you got to implement it. Then you got innovation coming out of that, but you got compliance, legal issues. So kind of some stall there. I won't say it's going to be a blocker forever, but if you go look at the bottoms up in the dorm room, the entrepreneurial activity, the developers, it's off the charts. They're intoxicated with this stuff. They're exploding with creativity. And so the question is, how do you meet it in the middle? What happens next? What's going to be the, how do you make this? How do we not lose momentum? Yeah. Between keeping the C-suite excited and incompetent, okay? And the developers keep voting. I don't think there's any danger of losing momentum. Exactly. I think that train left the station a long time ago. Who gets to the finish line first? Bottoms up or not down? I don't know if there's a finish line there. Yeah, I don't. It's a journey as out of what I hear. I bet you think we're going to get to an end at something. Yeah, no. There's a middle somewhere, I guess. I do want to point out, one of my favorite things on Reddit is the subreddit side project. And the reason being is that's developers, who guys who have a day job working at a tech company who go, I had this idea, bug me at night, I want to go build something, hammered it out. And what you're seeing is the quality of those side projects from three years ago was a roughly compiled Python app to now it's a full SaaS experience delivered on a cloud in five hours. Like they go in and build them. And so what I see these in the dorm rooms doing is they're creating new, I'll use a branded term, new Lego blocks of innovation that you can take and build on. They're not monolithic systems, they're components. And I look at this wave of creativity coming on and watch this in the marketing tools area. It's like there's a thousand little marketing tools. The productivity is off the charts. Yeah, a thousand of them. But they all plug into HubSpot or Salesforce or SalesCloud or Dynamics or whatever. They all hook in together. They're all built for integration. I think that's one of the dominant things in kind of the way we develop now is people building services on a cloud, build it with the intent, it's going to connect to a lot of things as opposed to be a silo. Awesome, I agree. I'll just add one more thing to it which is what Raghu talked about in his keynote today around the private AI, right? And I think it's a very, what you talked about in the dorm room is very relevant. When you talk about enterprise customers though, that's when your private AI piece will be a very, very relevant thing. For sure, yeah, yeah, yeah. Absolutely, and I think that's the platform piece that with VMware and the rest of its ecosystem that's been provided and to build on top of that with partners like Kindrel who can actually bring all of these things together with different cloud providers as well, is going to be key to success for us. The solutions for the C-suite. Absolutely, as well as the dorm room. The risk of compliance and the fear of F is going up. So in as much as we see a lot of creativity and the solutions that hook up into all these places and the data is transferring where you didn't expect it. Yeah, I mean, I love that earlier point about apps being up here and then the development of the data that's critical. And we even coined the term the data developer because we think a new class is going to be coming out of this trend which is coding data. And we've got platform engineering's booming. And so I love that fact that we got rid of the SRE word because that was more of a Google thing, but we have platform engineering now. It's like the version of SRE, but for enterprises. So you've got platform engineering becoming much more robust from a skill set. Yeah, AI and these data developers, the absolute is going to fall in place. Well, I want to go back to your earlier comment just for my final thought or kind of finishing thought, but top down or bottoms up. The answer I think is both, right? There's people right now that I think are developing things that can be leveraged at the enterprise level. And definitely an enterprise might take that as basically a private model and say, someone has really come out with a way to do something really transformative, leveraging a lot of technologies. And because of AI, it literally will scale. And open source allows people to contribute really fast and they can be in a dorm room or on a side hustle or whatever. Okay, that doesn't qualify as a final comment. I'll give you a model again. Well, you said two minutes. Technically, we're going to go around the horn. Ross, we'll start with you. We'll go around the horn. We'll end with Sunil control. Final statement. I think one of the things I, it's really one of the phrases we use at Oracle is clouds are built, not bought. And one of the things I like to celebrate is the rise of the builder, whether it's in the dorm room or enterprise, you see people now building systems with the design that it's really meant to be integrated. It's meant to have APIs. An API first development model has taken hold. That's so much better than the old ways we used to do stuff where closed, my app is an island, nothing connects to it. I just think we're on the verge of unlocking a significant amount of intellectual property that's been locked up in ways that we can't conceive yet. I think for us, we're really, if we focus on our goals, empowering every person at every organization to do more. And that's to do more with all of us collectively and leverage their history level, the investments that they've made up to where AI can build on what they've created and really go forward. So it's an extremely exciting time for all of us. You know, we're in that fourth stage again, the first stage, you know, the PC stage, then we had got to the web stage, then we got to the mobile stage. Now we're at the AI stage. We're just at the beginning of AI stage. It is extremely important for us to work as an ecosystem to build what's necessary to come to an agreement because we're going to be having different technology. We need to be coming to an agreement and we need to have policy and clients and security all baked into it because that's the only thing that'll take us to the next step of delivering customer success. Sunil, I'll take us home. You have to pull it all together for customers that can drill, take us home. Get real security, take it home. So the technology and the technologist in me is very excited about all the things you talked about and how every developer is more empowered, every business user is more empowered to produce, build applications, get access to data, come to insights. The challenge for a large enterprise is to ensure this is done in a risk-managed manner. And that which works, scales operationally and from a performance perspective, cost managed as we discussed earlier and pulling it all to that complexity together is a very exciting challenge and it's wonderful to have partners who are investing in helping it make it easier for the customers to achieve that. And MultiCloud is the best example of this but really these technologies everywhere but MultiCloud is really pulling it together and I'm really excited for this time of the industry on what we can do for our customers. Thank you so much. Gentlemen, thank you so much. We actually packed a lot of data in that interview. Absolutely. When we put that through our generative AI algorithm, I think we're going to have some highlights out of that. That's awesome. I hope so. Ross, Andrew, Monty, Sunil, thank you so much for sharing your perspective on this modernization, MultiCloud, SuperCloud world, really appreciate it coming on theCUBE. Thank you very much. I'm John Furrier. That's a wrap for day one on this set. We're going to do an analyst wrap up on the south of the set over there. You're watching the live CUBE coverage. Go to siliconangle.com for all the action. That's where the traffic is. CUBE.net where all the videos are and our community. Signing off here for this historic panel and VMware Explore.