VMware Explore 2025 | Sharad Kumar & James Bayer, FluidCloud

Clips
More from VMware Explore 2025

Sharad Kumar

Co-Founder & CEO

FluidCloud

James Bayer

Co-Founder & CPO

FluidCloud

play_circle_outline Introduction of theCUBE at VMware Explore 2025 with Paul Nashawaty and guests.

play_circle_outline Sharad Kumar and James Bayer discuss their contributions in supercloud and FluidCloud technology.

play_circle_outline FluidCloud enables smooth workload migration between different cloud providers and simplifies interoperability.

play_circle_outline The importance of reducing complexity and skill gaps in the developer community.

play_circle_outline FluidCloud offers transparency on costs, security, and performance across different cloud providers.

Info
Transcript

Sharad Kumar & James Bayer, FluidCloud

Sharad Kumar

Co-Founder & CEO FluidCloud

James Bayer

Co-Founder & CPO FluidCloud

At VMware Explore 2025 in Las Vegas, FluidCloud co-founders Sharad Kumar (CEO) and James Bayer (CPO) join theCUBE’s John Furrier and Paul Nashawaty to unpack how their company is rethinking multi-cloud strategy with a fresh take on workload portability. The discussion dives into the challenges enterprises face when workloads become locked into a single provider and how FluidCloud is enabling true cloud fluidity by scanning, mapping and translating workloads across VMware, AWS, Azure, GCP and beyond.

Kumar and Bayer explain how FluidCloud’s recent eme... Read more

explore Keep Exploring

What event is theCUBE covering and who are the featured guests? add

What is the concept of supercloud and its relevance in multi-cloud environments? add

What is FluidCloud's perspective on workload mobility and the options available for managing workloads in the cloud? add

What are the major barriers organizations face in the modernization of applications and the movement of applications across different environments? add

What features does the tool offer for analyzing cloud provider costs and decision-making? add

bolt Powered by CUBE AI

Sharad Kumar & James Bayer, FluidCloud

search

>> Welcome back, everyone. We're theCUBE, we're live here in Las Vegas for VMware Explore 2025. This is the 16th consecutive VMware annual event usually should call VMworld. Of course, theCUBE, this is the longest-standing show. VM now part of Broadcom, certainly has changed. I'm here with Paul Nashawaty, leading analyst for AppDev, cloud, cloud native. You see him at KubeCon, all over the place, Kubernetes, you name it, he's on top of it. Two great guests. We've got two founders in here who are making waves in the industry by abstracting away and building a supercloud. Of course, VMware got a lot behind, cross-cloud, multi-cloud, this is now starting to see that happen. Sharad Kumar, co-founder and CEO; James Bayer, co-founder and chief product officer. Sharad, great to see you.

>> Good to see you.

>> James, thanks for coming on theCUBE.

>> Absolutely, great to be here.

>> Welcome to theCUBE. CUBE alumni, you now made the list.

>> All right, I'm happy to be here.

>> James, I mentioned supercloud because in 2011, I mean 2021 I should say, Dave Vellante came back from re:invent, I'm like, "Dave." The cloud, the top SaaS players, Databricks, Snowflake at that time had the Data Cloud. They weren't calling it the data SaaS product, they were calling it the Data Cloud, but they had no cloud, and then they were starting to go to Azure. Databricks obviously was on AWS, and then had to go to Azure because customers had a lot of different products from Microsoft, for instance. We saw the need for true multi-cloud where you can pick your best cloud for things you wanted, but they weren't really talking to each other at the data layer. We saw the wave of coming around, what is that abstraction between the different clouds? Because lingua franca cloud on Amazon was not the same set of primitives on Azure, so it didn't take a rocket size to figure out you need some sort of translation or abstraction.

>> That's fair.

>> We called it supercloud. Got a lot of people throwing tomatoes at us, calling it hype. So we put out a working document, Charles Fitzgerald at Platformo, he hated it, but we said, "Hey, contribute." So that started the supercloud, our narrative, but that was really early, but now you're starting to see it, you guys are doing it, so I want you to explain the product. What does FluidCloud do? Because you guys are getting traction in an area around both migrations but also interoperability between environments. So if they've got a private cloud like VMware VCF 9, they're trimming down and focusing on a core set of companies that map to their ecosystem, which is now large players, not large numbers. Azure's here, Google's here, AWS is here, and then you've got this new developer environment going crazy. So what is the FluidCloud product?

>> Great, great question. FluidCloud is all about the name fluid. We believe that your workload should be fluid, and you should be able to move them to the cloud provider that you want and move them to. For far too long in this dynamic that you're describing, basically people have put their workloads in one cloud and they've just stuck there and they stay there for 10 years or maybe even longer. One of the things that's exciting about this conference is, as you mentioned, we see the big three hyperscalers are here. We're actually here with a partner, OVH, and a neocloud trying to offer other options as well. And so the number of options is really increasing, but if you can't get your workload out of one cloud to another cloud, then you're stuck. And so we believe that you should be able to do that and our technology enables that. Basically, we scan the cloud that you're on today, we build up an inventory of what you have across all the providers, and then we translate that compute network storage and similar services to equivalent but different services on other clouds.

>> Sharad, you also came out of stealth.

>> Yes.

>> What data did you guys reveal and what's been some of the traction? Give us a little taste of the timing of stealth coming out of stealth and commercial opportunities.

>> Yeah, we came out of stealth almost now a month, 29th of last month, and we picked up 8.1 million from Unusual Ventures. And currently, from the opportunity perspective, we are getting ton of opportunities, more from the people who want to move from on-prem to cloud, cloud to cloud. Those are the opportunities we're looking at.

>> What's the demand curve look like? Give us a taste, people lining up, what's the velocity of deals? You guys overloaded? What's the capacity?

>> Yes. No, absolutely. Multi-cloud is primarily an enterprise problem at this point in time. Most of the enterprise are stuck with that. So most of the demand, which is coming at this point in time is coming from primarily the Fortune 2000. And that's where we are able to see that this demand is just ramping up at this point in time.

Paul Nashawaty

>> Yeah, that demand is really definitely growing. What we're seeing in our recent research, we just had a nice set of research come in from the field, amplifies everything you talk about in your value proposition. We were talking before we got on stage here and talk about this. 94% of the respondents in our recent AppDev research indicates that they're running on two or more clouds, right? 65% say they're running on four or more clouds. The complexities out there, to your point, James, you were talking about workloads being implemented on a cloud, it's set it and forget it in most environments and because it's so rigid. So what I like, what I'm hearing, and obviously congrats on the coming out of stealth, that's really awesome, but there's definitely a need in the market for portability. We actually see in our same research that 20% of organizations want application portability in their environment. 67% say it's very critical. There's a lot of data to back what you're talking about specifically from-

>> Go ahead.

>> One thing which I definitely want to add to that is all the people who are into this multi-cloud model using three clouds, two clouds, or four clouds, majority, like 90%, 95% of them, when they are in multi-cloud model, majority of their workload is on one cloud, and very small workloads are on the other cloud providers. What's happening is that they're not able to rebalance this portfolio. The way it happens always with the new cloud provider is they have a couple of good credits which they offer and they don't move the whole stack. They just now the new stack gets start building on top of it. But what we are envisioning at this point in time is when you have to balance your portfolio, you can do that without this hassle of having 90% on one cloud and the rest 10, you can distribute it among rest three or two, whatever it could be. That dynamics will change.

>> Well Paul, I want to ask you a question because I had a chance to interview Sharad at the NYSE. AppDev is developers, so they care about deploying, right? So you've got the Kubernetes world going beautifully.

Paul Nashawaty

>> Yep.

>> As AI comes in and the cloud pressure with generative AI, the on-prem action's hot.

Paul Nashawaty

>> Yes.

>> VMware loves that because they can have a private cloud.

Paul Nashawaty

>> Yes.

>> Okay. But at the same time, there's half the audience of VMware here are migrating, right? So you have that problem. What does that mean for the developer community and what does your research tell you on the dynamic between, because the pressure to build devs on the dev side, is that an all-time high? I mean, I just saw Replit just 10X their revenue, divide coatings happening. What does that mean for all this?

Paul Nashawaty

>> Well, John, it's a great question, and when we look at it, what you're talking about from FluidCloud's value proposition, it really plays into where the future of AppDev's going. My practice focuses on past, present, and future of applications. Modernization's a treadmill that's continuous, you continuously run on it, right? But the major barriers that organizations run into is basically complexity and skill gap issues. Those are two major things, on-prem, colos, wedge locations, public and private clouds, all of it is a challenge. With this migration or movement of applications, you're going to see applications going burst capacities to different areas. They might be moving to the cloud, they may be repatriate back on-prem. There's the need to have these different locations. The other thing, John, to bring up as well is application developers are no longer going to be the professional developers. We're seeing the growth of applications. At edge locations alone, we're seeing 500 to a thousand AI workloads, production workloads at edge locations alone. So if that's happening, that means the lines of businesses are actually starting to be the developers. There needs to be this little bit more transparency, less complexity, and reduce that skill gap issue as well. And that's what I really love about your value proposition.

>> Yeah.

>> Yeah, absolutely. I think to this point, we see AI as something that's de-thawing this stagnant cloud strategies because for the first time, you weren't able to get everything you want just from your hyperscaler that you might be using because the hyperscaler you chose, let's say seven years ago when you first started your public cloud lab environment, then it became like, "Okay, let's do a POC and you scale it up," all of a sudden you have seven, eight figure cloud contract, but you can, for the most part, the last, call it seven to 10 years, get everything you need from that one hyperscaler. But now, with AI services coming in, that's probably not going to be, the decision you made seven to 10 years ago on that provider probably doesn't match the best AI services available on the market. So now you're introducing new providers, adding complexity, and so there's going to be a new set of tools and services needed to manage that.

Paul Nashawaty

>> Well, and also the demands now on FinOps is a big factor now. It was a big factor then this little thing called the AI came in and FinOps went out the window for a minute there, but now sustainability and FinOps is coming back in as a main player, so I think you're right on with that.

>> Let me ask about the cost piece because this comes up a lot. One of the things I hear is the costs on skill development and complexity management is just another transformation project.

>> Absolutely.

>> So they go, "Okay." Now, the other thing that we were doing with the supercloud analysis that we did at theCUBE Research was looked at this abstraction layer that you guys now have built, and are continuing to build. We call it a super pass. Now, you were at Cloud Foundry.

>> Yes.

>> You saw that game. Won't comment on the success ratio of Cloud Foundry, but you're still seeing Salesforce has Heroku, right? You still have developer environments. People are going to have their own little sets. Also, you worked at HashiCorp.

>> Yep.

>> Okay. Very successful in infrastructure as code. Okay. So now if you take those lessons learned and apply it to today, how do you look at today's pass layer? Because all the infrastructure is continuing to be retooled. You got GPU clusters coming out, hot chips is going on in Stanford right now, more capabilities coming, which is going, and now that Kubernetes is standardized, that's a good thing.

>> Yes.

>> Okay. Infrastructure code has not changed, that's a good thing.

>> That's a good thing.

>> What is that next super pass or DevOps, DevSecOps enablement look like in your mind?

>> Yeah, well, I told the story a little bit earlier around the startup founder I talked with last week that is assembling their new cloud platform as they go. They took Vercel for the app tier. They took Supabase as one of these databases of service for best in class databases. They didn't go to RDS, they could have, but they went to Supabase because it has more developer friendliness to the point of the productivity for the end user, right? And then they took S3 for the object storage. And you've seen even CloudFlare, I don't think they're here at the conference, but they introduced a container as a service they're getting into past. I think increasingly we're going to see this big array of cloud services, and the thing is you want to be able to have that visibility over all the services you're using as well as for the ones that make sense for compute and storage. You want to be able to move those around because sometimes this workload is very data intensive. Well, some of the providers charge you incredibly high data transfer fees, and other providers don't charge you very high fees on that data transfer. That workload is best suited for a place that has low data egress fees. We think people just need to have that flexibility to move things around, but that starts with the visibility so you can see everything that's there and then optimize from there.

>> On the cost side, what's your reaction to the reaction people have about cost? Well, if I move off VMware and go to say, another abstraction layer, what's the cost angle there? The first thing that goes, "I got to learn FinOps." Like okay, do they or don't they?

>> I think one of the things, and which we are very proud of, is one of the features of our tool is you can have, once we do the inventory scan, we get you all the all of the cost details of your current cloud provider, and we also map it exactly to the cloud provider you're planning to move or the infrastructure you're planning to move. And also, the way we are going to build this as we are progressing towards what if I would have moved six months back to a particular provider, how much I would have saved? All of this analysis today is-

>> So you're doing FinOps, kind of.

>> It's part of it.

>> Paul could comment on what that means, but.

>> Yes. It's part of it. So that motivates people to rethink about their whole portfolio. And so basically, it's the RGB, the way we think is the RGB of infrastructure. RGB would be like the cost is one part of it. The security is one part of it, the performance is one part of it. If the user gets an absolute visibility of all these three, they'll be able to make better decision for their cloud provides. That becomes the RGB and all of the colors comes out of this.

Paul Nashawaty

>> One of the things I do want to point out that I find very intriguing is a lot of the hyperscalers that are here, they'll basically say, "Well, we're going to give you the cost factors as the Happy Meal toy that comes with it." So they know they're going to tell you, "Here's what comes with it," as their tech stack. But what you're doing is you're saying, "Okay, it's not just one cloud." You're seeing it holistically across the entire ecosystem and you're able to put the onus back on the practitioner. But what's interesting, a lot of the DevOps teams will be like, "Look," they're going to over-provision because they just need to get something done. They don't own the budget. But now that there's pressures on the budget, they're going, "Okay, well we need to get this done, but we need visibility now." They don't want to have individual tools. As we were talking about previously. We were talking about the fact that if you know a developer that knows GCP, AWS, Azure, well first, you pay them anything that they want because you'll never find that person again. But second, it's impossible to find somebody like that. So what you're doing is really, really powerful.

>> I think the demo that we show to people that gets them really excited is we take a full stack application, compute network and storage and maybe even some security services like IAM and things like that, and we have it running on one cloud provider, and in a few minutes we show them that same stack running on the other cloud provider, and they get to choose which one you want.

>> It's not premeditated on their side, you just do it for them in the call.

>> Yep, that's right.

>> That's a good demo.

Paul Nashawaty

>> That doesn't mean that they don't have to learn the skills of the destination cloud they're going to be on, but it accelerates it because they can use the knowledge they already have on the existing cloud. Like, "Oh, this is how that maps to the concepts on this other cloud. Maybe this isn't so hard, it's not as intimidating."

>> You guys call that cloning.

>> That's right.

>> You clone the service set, and then look at what it's going to look like so they can see full abstraction of that service that runs there. So I have two observations and use cases. I want to get both of your reactions, and I'll start with a little lead in, preamble. So the top sponsors of this year's VMware Explore is, again, if you know VMware, you know what it looked like five, six years ago or hundreds of sponsors. It looked like it was so small the logos couldn't fit on the sponsor. Top sponsors AWS, Google Cloud, Microsoft. Not Azure, Microsoft, which includes Azure, probably some other things. Intel, Dell Technologies, IBM, NetApp, VMUG, of course, AMD, Pure Storage, Rubrik, World Wide Technologies, WWT, and then AMD and a bunch of others. That tells me the ecosystem has moved from a thousand flowers blooming to heavy hitters. So that tells me two things. VMware is letting go their low end, focusing on the high end, private cloud, and they want their private cloud use cases to connect to other environments, so they could use a cloning service to understand interoperability. And two, all the other customers who were leaving, and they're probably planned for it intentionally. So two use cases, not the big customers. They're going private cloud and VCF, they're already in, they don't need to sell. It's just, it's plumbing to them. They're using it. The estate's getting manageable, it's bigger. VCF 9's heavier, fatter, bigger, more bulkier, more bang for the buck they say. Okay, great. I got a private cloud. I want to work with AWS, Google, Microsoft, Dell, IBM. That goes on and on. Okay. So that's one use case. And then the migration, a lot of people are like, "Okay, I'm gone." Complexity and skills gap, that's what his research shows. Okay, check that fits. Okay, tell me how it works. Take me through those use cases. How does FluidCloud handle the new modern VMware, Broadcom, VMware? Because clearly their strategy is sound, we're going all in on private cloud.

>> Yes.

>> Enterprise cloud basically, and now the rest, just fend for yourself.

>> James, you want to?

>> Okay, sure. Let's tackle the VMware piece first since we're at the conference. We support all of the VMware environments, including NSXD, which is something we hear often about for people that, let's just put one of the buckets there. You said the smaller customers that don't feel like-

>> Yeah, I'm gone, price increase, I'm out of here.

>> Yeah, whatever the situation is, they've made a decision. One of the things we've heard from them is that when they start migrating, the NSX team mapping in particular, because you can have firewalls, virtual networks, subnets, load balancers, all this kind of stuff wrapped up in that, that can take a long time to translate. We've automated that piece. That's something that can accelerate on the technology portion of that. We talked a little bit about the skills portion, and then some of the customers we're speaking with have up to 10,000 VMs or even more to migrate. Those are things that we're trying to accelerate that piece too. That's on the VMware side, but then you talked about these three hyperscalers here. Well, if you're going to be doing VMware on AWS or Azure or GCP, you're also probably going to be using those public cloud services as well. So now you have a multi-cloud estate that managed. We give you the visibility, and the cost comparisons to make choices when you're trying to decide which.

>> You help the VMware customers going all in on VCF 9 because you're saying, "Hey, we have clone and we can figure out which primitives to work with on AWS based on whatever your identity might be or whatever your dev environment looks like." On the migration, I'm more interested in, because that's obviously very current in terms of it gets a lot of chatter on X. Tons of discussion. I've heard both sides of the coin. It's hard as hell, it's expensive, I'm just going to give in. Or no, I am going all in moving. What's the scope of the magnitude of that migration?

Paul Nashawaty

>> And actually, to echo one more piece on what John's talking about, it's not just about the technology here. It's about the learning, the new tech. So that's something in your response, I would love to hear how you're addressing the tech and the skill issue as well.

>> Yeah.

>> Go for it.

>> Okay, so let's take a traditional migration. Let's say this 10,000 VM customer that we're talking about. It's a real customer. We're working with the state-of-the-art for them up until this point is to hire an SI and to work for many, many months, workload by workload to get this 10,000 VM. Yeah.

Paul Nashawaty

>> Super expensive.

>> Every workload of let's say a hundred VMs, out of that 10,000 could cost you six figures and take six weeks. So you're chopping this a little bit at a time. And in this case, they have a certain less than one year before their renewal. And some of the terms on these renewals, they want you to do a three-year renewal. So it's not like, "Oh, well I have one more year of the contract, I actually have three." So they're strongly motivated on timeline. They're not going to make it though if they're doing that one workload with the SIs.

>> Too slow.

>> It's just too slow.

>> And by the way, their operations are changing, they have demands.

>> Yes.

>> They're trying to swap out the airplanes, put a new airplane at 34,000 feet. Build a bridge between two airplanes at 35,000 feet.

>> Yes.

>> All the best.

>> The mapping that we were talking about earlier does accelerate on the technology piece. The skillset piece is going to be something, it will be a challenge. We can't do the matrix thing and inject all the cloud-native into a traditional sysadmin.

>> Not yet.

>> Not yet.

Paul Nashawaty

>> At least.

>> Not yet, Neuralink.

>> But we think that we can accelerate, as I mentioned before, and lower the time that it takes. Maybe we can't get you to a level 10 on a public cloud from if you're a VMware sysadmin, but maybe we can get you to a level three within a few weeks. That makes the journey to going to level five, level seven, whatever, more approachable. That's how we're looking at it there. And this team actually has AWS already, and so they do have people in their organization that can learn from.

>> Scope that migration timing, I don't have any clue because in my mind, I'm with you on the other one, it's going to say months and months and months. Not everyone's going to slice the salami each by piece by piece. They just want to put bulk, the whole thing. If I say I'm all in, here's a check. What's happening? How fast does it go?

>> That's what I talked about this demo earlier. We can do the technology pieces in minutes. You're not going to do it in actual minutes. You're going to do it in probably a period of days and weeks. But we believe we can dramatically shrink that timeline from months to at least several weeks on that. Let's call it a hundred VMs at a time or something like that. That's that piece. There's obviously some skills transfer, some operational readiness, make sure the DR works. There's some of these process things that you're going to have to practice in the new regime compared to the traditional one. And that takes some time, you can't speed up. But going from months to week sounds pretty good to me.

>> Yeah, that's a much better. Any comments on as-

>> Oh no, absolutely. One of the very important portion of this whole migration journey is when you're talking to an SI, even SI is clueless at this point in time from the resource perspective. So they're going to give a huge shout out of, "Oh, I need a hundred of people who have the skill set." Can they come on this? Another thing on the skill set is even the people know that skill set, they have to write the whole Terraform all the way back again. The way we are doing it is we have created, we get the inventory and then we have the mapping. Through the mapping, we create the Terraform. And through that Terraform, the deployment could happen. Now you don't have to write the Terraform all the way back again now for AWS or GCP or Azure or any of the cloud, which you're thinking of. Our philosophy is very simple, we'll meet the customer where they are. We'll not ask them to reinvent their journey.

Paul Nashawaty

>> That is awesome. And actually, I would love to have part two of this discussion on our AppDevANGLE podcast because we can go really into the deep dive and the technology because I really want to talk about those workflows and the automation. Because everything you were both talking about, I'm thinking AI automation is there and it's going to help with this whole solution.

Paul Nashawaty

>> Well, it's an interesting one. We wasted almost four months initially. We started this company last year, August 8th, and we almost took four months just working through the AI.

>> A lot of founders know the exact date, probably know where they were. Five o'clock on the 8th.

>> We were at the Black Hat Summit at Vegas, so exactly around this place, we founded the company. The problem was almost four months, we tried doing this stuff using AI. It doesn't work that way, so then we started going through the hard-

>> AI could help on the skills gap down the road as you guys get data and workflows. But it's not a silver bullet today.

>> No. There's a lot of hallucinations where it would say, here's what the translation from this cloud provider to that one is. And then you'd look and they'd say that API doesn't even exist that it's telling me to use. So we're like, "Okay, this is not going to work."

>> You're getting hallucinations.

>> Approach it a different way.

>> And every single time, and we are trying to deploy the infrastructure, it has to be 100% accurate. So you do not get that.

>> Well guys, take this to the AppDevANGLE, his podcast. Go do a deep dive on that for sure. Great to see you. Final point, put a plug in for the company, just out of stealth, just did your first round. So you're young, you're out of the gate, get traction, what are you looking to do? What's your goals? What are some of the things you're working on?

>> Yeah, no, absolutely. We have these enough leads at this point in time, which we are working with on the POC side of the world. We just want, so instead of having 1,000 people who, like you, Brian Chesky said, you need 100 which loves you. So we are looking for that narrow group basically, which loves us, and then they become our evangelizers. The whole point of that is now we just want to make sure that we have all these customers, we are able to move them as fast as possible before their timelines. That's one. And then we have the multiple use cases beyond the migrations that you're going to discuss on your podcast.

>> And by the way, being global with it is going to be a big VMware strategy. Sovereign cloud should be a nice pop for VMware because they're going to want to put that out there and you're going to need to talk to other clouds.

>> Absolutely.

>> So again, that's going to be good for you guys.

>> And that's one of the reason we partnered with OVHcloud because OVH is primarily a France-based cloud provider, and then we have a US-based too.

>> Guys, thank you so much for coming on theCUBE. Appreciate it, coming in here at VMware's event.

>> Yep, thank you so much. Really appreciate it.

>> Migrations, connecting clouds, multi-cloud, supercloud, whatever they call a super-pass layer. What's happening is an abstraction of resources, mapping them in real-time with software will create faster accelerated connectivity, connective tissue, and ultimately a better developer experience. I'm John Furrier, Paul Nashawaty here at VMworld, VMware Explore 2025.