Cloud, Edge & AI: What CXOs Are Planning Next (with Solutions II)

Bruce Kornfeld

Welcome to PodMagic, real conversations about solving real IT challenges. My name is Bruce Kornfeld. I’m the Chief Product Officer at StorMagic. And we’re always exploring how simple, reliable technology can benefit you and the people that you serve, no matter what you do. Whether you’re running branch offices, retail stores, or just supporting customers on the front lines. My goal is to bring interesting guests, deliver value and have some fun along the way. For today’s episode, it’s going to be a little bit of a CXO round table in that I am the Chief Product Officer and I have two gentlemen here that also have C in their titles. The first one is Jeff Guenthner, the Field CTO from Solutions II, as well as John Wondolowski, the Chief Technology Officer of Solutions II as well. Thanks for joining gentlemen. I think what we’ll do is have you introduce yourselves. What do you do? What’s your area of focus? And a little bit about your background. Let’s start with John.

John Wondolowski

As you mentioned, Chief Technology Officer for Solutions II. We have people around the United States. We have a few verticals that we’ll probably get into where we have some depth of capability. We’re a managed service provider, but we’re a solutions company, value-added reseller, system integrator. I’ve been with them for about seven years. I spent most of my career in enterprise IT as an enterprise IT leader. Then I went to a value added reseller about 13 years ago and tried to lead their transformation into the new cloud era, but there’s always the next transformation and the next digital disruption. Then Solutions II bought that company and I’ve been with them ever since kind of like Jeff.

We’re all client facing and our job is to make sure that our client facing time is productive and helps our brand. Specifically one of the things I do is I build out our ecosystems, could be client advisory boards in our verticals, which are casino gaming and public safety, or in the broader general IT sense. We lead a CIO round table each month, for 13 years, and we lead a chief information security officer round table each month for 13 years. Each of those groups has over 100 members, no cost, no sales. We just facilitate the round tables and bring in subject matter expertise. It’s about networking. It’s about learning from each other, and it’s about us learning from them as well.

Bruce Kornfeld

Well, I’m excited about our little roundtable here. A bunch of tech guys talking tech. So that’s what it’s all about. Jeff, why don’t you do a quick intro as well?

Jeff Guenthner

Similar to John, I was working for the company that acquired Solutions II. That’s how we became besties. I’ve been with Solutions II for 20 years, working in a field role of some kind. In my role now as field CTO I do a lot of things. I do a lot of client engagement like John’s talking about. I specifically work primarily in our data center practice. As John mentioned, we have two other verticals, gaming and public safety. I dabble in those when I’m needed. John does a lot more work with them than I do. I spend a lot of time talking to what we call our data center practice, which effectively is anything that’s not casino gaming, commercial, tribal or public safety in that space. There’s a wide swath of clients that I talk to in healthcare and consumer package goods manufacturing, retail. It’s just this nice long tail effect for us. John and I look at our ecosystems, our portfolios and decide go to market strategy for managed services and things that we sell and building outcomes that give us competitive advantage in the marketplace. But by and large on a day-to-day basis, I’m usually interacting with a client in some form or fashion.

John Wondolowski

Probably most importantly for the last 10 plus years, Jeff and I have been co-managing a fantasy football team and our team is the winningest franchise in the history of that league.

Bruce Kornfeld

You guys must know what you’re doing. That’s awesome. Even though you’re both client facing, I like the fact that you’re likely not the people asking for purchase orders. You’re the technologists that are helping the process. That’s what it’s all about.

Jeff Guenthner

I’m glad our CEO is not here because our CEO would say everybody sells. It’s just a matter of, we deal in outcomes.

John Wondolowski

For better or worse, I’ve never seen a purchase order to Solutions II. I’m not there when that happens.

Bruce Kornfeld

You cover all sorts of verticals, and John mentioned the word cloud already, but what do you see from your end users, from your clients, in terms of cloud versus edge? Where are they thinking about putting infrastructure these days? Is it “I’m all cloud, I’m all edge.” Is there a lot of debate in between? Let’s open with a good conversation about cloud versus edge.

Jeff Guenthner

I think all our clients have a foot in both camps. There was an era where everybody was trying to get to cloud because it was the thing, as you alluded to John earlier, there’s always some kind of thing going on. I remember cloud, big data, digital transformation. Now we’re in AI. There’s always going to be something going on. I think that now a lot of our clients are looking at their cloud workloads and really trying to decide what’s best for them right, because now it’s about operating margin. We see workloads that should remain in the cloud because they provide nice business agility. And we see workloads that are probably better off coming back into a data center of some kind, whether that’s their own or a co-lo facility, because they can just control the operating costs much more elegantly. The workload is known and it’s not really elastic. From an edge perspective, that really depends. Our clients that are doing, consumer goods manufacturing, or more specifically retail, I think those folks are starting to do things at the edge, especially around analytics, inventory supply management. Some have pharmacy apps because they have a pharmacy in their stores that they need to have access to. Some point of sale vendors say they’re cloud, but they still require a gateway in the store, some are full cloud. Like this gamut depending on the retailer and what they want to do. I think we see both is what I would say and it just depends.

John Wondolowski

I’d say that edge, which is not really well defined yet, is defined a bit differently depending upon industry. Would you agree with that?

Jeff Guenthner

Yeah, I do. Absolutely. Yeah.

John Wondolowski

I’m very fortunate, we have hundreds of clients and then we’ve got these round tables each month with another hundred plus in each of the round tables. I get to hear what’s top of mind and the top of mind in every month, we do in-meeting pollings. And so we get an actual quantitative view of what’s top of mind. Top of mind right now is AI adoption across industry. Our round tables tend to skew to mid-market, small enterprise, depending on your definition of that, but they’re across all different industries and the AI adoption is the wild card for edge and cloud. I think it’s going to somewhat alter that balance between cloud and edge, because if we go with AI adoption it’s really about data. Data gravity and the speed of real time decisions.

Bruce Kornfeld

Can either of you think of an example where you’re working with a client and they made a decision that said “we need to run on prem at our small locations”? They have multiple small locations in addition to a data center or a cloud somewhere else and they’ve made a decision. They can’t run their applications in the cloud. They have to run it locally, or the other way around.

John Wondolowski

I definitely can because I was in the conversations where this multi-billion dollar client in food manufacturing, they got the food and then they packaged it, they said, “we’re moving from our Oracle ERP on-premise and we’re going to go all cloud. We’re going all cloud.” We planned that out for them, helped them plan that out through a few workshops and got to the point where it was pretty clear. Some of it was their locations and the network capability at those locations. Some of it was related to floor manufacturing processing and that type of thing. And they reached the decision on their own, with us, that no, we need at a minimum, edge, if not completely on-premise to a certain extent.

We’ll go as cloud and to go a little bit further, when the pandemic hit we were talking to clients over and over again, different ones about the need to go all in on cloud. For those that weren’t born in the cloud. I can count, out of 40 clients that we talked to, how many were able to go 98% cloud. Each of them had a use case where they said, “that’s not going to work, we need on-premise capability”. Jeff, you remember who I’m talking about?

Jeff Guenthner

Yeah. That specific client has an application that they’ve built that’s custom for them and their manufacturing processes that they run at each location. They have a lot of locations. They had to at least keep that on-premise for obvious reasons. Like the transition to cloud, a lot of it for us has been like big ERP systems moving, getting broken up and moving to SaaS platforms like a Workday or NetSuite or SAP in the cloud, et cetera. You used to see a lot of infrastructure around those applications and the business has just simply decided, “no, we’re going to pump that to somebody else, and it’s going to be cloud”. We still see a lot of infrastructure as service with the applications that clients maintain for themselves. And then those analytics applications, they might collect data from the edge and bring them in to a main processing source, but they’re still collecting data at the edge in some way.

Bruce Kornfeld

What’s interesting for me listening to these stories, is that StorMagic is a software provider in the infrastructure space. We typically aren’t at that point in the conversation. We get pulled in by people like you and end users when they’ve made their decision that says, “we’re building on-prem edge infrastructure” and then we become a choice. Hearing the upfront conversation about how it goes, sometimes right or wrong is quite interesting for us because we’re usually not at the table for that level of conversation.

Let’s move to another topic. Let’s talk about hyperconverged. The whole concept of, instead of massive server farms and storage, just converge everything, one server, two servers. What are you seeing in the vertical markets? Is this something that has passed its prime? Do you still see end users focused on it? Where do you see hyperconverged infrastructure in your customer base?

Jeff Guenthner

I think everybody’s pissed off at Broadcom and wants to leave VMware. There’s a lot of business to be had in that space. I think you fit that space very well in that you offer an elegantly packaged solution that I can run on at least two servers for consistency. I can have this very lightweight witness. I can either run it in my environment, or I can run it in the cloud as a service, which is very inexpensive as well. And it just works, right. It was designed for that. Our clients love it. It’s fantastic. I equate what you’re doing to what IBM did with a mainframe or an IBM. I write the AS 400, the thing just runs and it’s reliable. It’s very cost-effective and we love the fact that you guys are making your own hypervisor now, which is how I fold back in the Broadcom conversation, because it’s going to be a lot easier for clients to make that decision. Because right now they’re struggling. “Do I stay on vSAN? Do I go to Microsoft with Hyper-V? Do I do something with Proxmox? Do I just go a completely separate direction and do open shift virtualization because I want to go to containers eventually?” These are like application level, business function level decisions that get made that ultimately drive the architecture to the client. That’s my perspective, at least. John, what do you think?

John Wondolowski

I completely agree with you. We’ve got clients in our casino gaming industry that you wouldn’t think were casinos. They’re like hundreds of gas stations or convenience stores that happen to be located in a footprint where they can have slot machines. For a wide variety of reasons, there is momentum for hyperconverged infrastructure on the edge, in use cases for casino gaming. And of course, Jeff mentioned the Broadcom thing and that has really created, a big interest from the IT leaders on, what’s my solution. The most important thing right now, because the runway that they have, is that it’s going to take a lot of change capital and is it really that much less expensive? The ROI needs to play out. But then in the end, some large percentage of our clients would, this is an informed guess of 50 to 60 % are saying they don’t have the ability to change right now. I’m holding my nose and signing up one more round, which means that they’re not done with the migration. They’re not done with changing their edge devices. They’re just not ready for it right now. That whole VMware in edge use cases, it doesn’t have a lot of legs, especially in the SMB and the small enterprise, and even really in the enterprise. It’s just because that doesn’t seem to be the target demographic that Broadcom’s going after. So there’s going to be more fallout. The one thing I mentioned earlier with that food processor client is, they do come to us and say, “now what do I do?” And one of the real drivers strategically that they need to solve for is the data. We’re going to replicate data across places. What’s the role of the cloud in that? Because one of the things that we found from our round tables and our clients is the sprawl of data and the data silos has gotten to a point where it is hamstringing AI adoption or really any other data or technology transformation.

Bruce Kornfeld

I want to talk about the whole VMware thing, which is love and more hate. Hock Tan and Broadcom. They’re brilliant. What they’re doing there from a business perspective, their company’s thriving. Stock price keeps going up. Their revenues and profits are record high. They have some smart people there. And what they’ve done is they’ve created a whole ecosystem, a whole opportunity for software providers like ourselves to go fill the gap of all of the small customers that they’re leaving behind, as well as some of the big ones that can’t afford VCF at their edge site. It’s great though, because now you guys have customers coming to you with, what do we do? How do we solve this? It’s been great for the industry. It’s frustrating for customers, but it’s just created some real interesting things going on.

One other question I had for you is, do you hear customers talking about getting support from third parties for VMware. Is that real? I’ve heard a little bit about it. Have you had customers go that way? Does it work out well?

Jeff Guenthner

Absolutely, it’s very real. It does. It works best with clients that obviously have perpetual licenses because they’ve signed those agreements and those are hard to litigate. I know Broadcom has sent system deceased letters and still continues to. But our clients are looking at third party support where they have perpetual licenses to get that protection, because they’re looking at the change capital is going to take and the transition costs. Maybe they’ll sign up, they’ll get to a state on version eight of vSphere, get all of the updates, get everything they need, and then transition to a third party, set a one to three year clock, and then start working out what the strategy is going to be for that transition, knowing that they’ve got a self-imposed deadline in three years. They could certainly re-up the support if they want to, but it’s a way for them to force that change.

Bruce Kornfeld

Clearly, every provider, every end user has to be at least thinking about what is our AI strategy? What do we do? What are you seeing really happening now with AI and how is it impacting customers’ decisions about infrastructure? At the edge for smaller sites, is it really starting to happen and what kind of decision making is it driving?

John Wondolowski

Jeff, I’m going to set you up on this so that you can hit it out of the park. I’ve mentioned the roundtables that we do. Three years ago, there was no chatGPT. 18 months ago, I pulled all of our roundtable numbers, Mid-Market, small enterprise, and ask them about their AI status and I used three terms. Are you active? Meaning that you’re experimenting. Are you operational? Meaning you have some stuff in production. Do you have substantial AI across the organization? And are you doing nothing? Just last week active 70%, experimenting. Operational 20%, in production. Substantial AI across the enterprise 5%. If you did the math with me, that’s 95% that are at least experimenting. I asked that question 18 months ago, exact same question. It was 25% experimenting, 5% operational, 0% substantial across the enterprise. There’s impact, but is everybody making the right decision? Everybody’s not making the right decision. Are people confident that they’re making the right decision? Data management, number one concern for IT leaders today, but everybody defines data management differently. Jeff, edge data management, what do you see?

Jeff Guenthner

There’s a commercial that’s on TV that I’ve seen quite a lot. It’s from another vendor that is an enterprise vendor, but there’s this interaction between these two individuals to talk about AI. One individual, clearly the customer is like, “well, yeah, we have data and we know data is all over the place”, the AI is using the data and the other person’s like “but you’re not getting any results, you don’t understand”. From our perspective, one of the things we tell our clients, even with co-pilot usage is, we approach it from a security and data perspective. You’ve got to understand what data is out there, what exists, where it exists and who has access to it. Because when you start giving access to co-pilot, it’s going to tell you anything you want. From an edge perspective, a lot of the stuff we see, even before chatGPT, same thing that Google search does, same thing that chatGPT does behind the scenes, it’s a massive machine learning algorithm inside of a neural network. It’s a little complex, but we were doing predictions for anomaly detection well before chatGPT came out. We see a lot of our clients are still doing that. They’re running some kind of algorithm to collect data for some reason and then send it to a central source to be further analyzed, for analytics, servicing of parts, you name it. We do see a lot of that happening at the edge. That’s a very common occurrence for our clients. And I would say that it’s predominantly the largest AI use case in our world. Inventory, fill anomaly detection on equipment would be the two largest. Then a lot of other clients are experimenting with “what can I create with my data that’s AI led or AI driven that I can then charge for as a service. How can I create more value for myself as a corporation, but also for my clients?” That’s typically what we see.

Bruce Kornfeld

Is the big thing here about organizing it, protecting it, securing it? Do you see any of this driving different infrastructure requirements, in particular more storage per site, more CPU needed, memory? What about GPUs? Are you getting requirements for that? Let’s talk about infrastructure being driven by AI.

Jeff Guenthner

Yeah, we have a client in the industrial space, they have an edge server that has a decent amount of CPU. We’re using the AMD CPU. It’s got a beefy memory requirement. The storage is multi-terabyte. It does have a GPU on it and they use it for some of the use cases I’ve described to you and others for their clients and their deployments. We’re absolutely seeing that work at the edge. If you can write software code or you have a vendor that you can work with that is in the AI space that has that capability, having a GPU on an edge server is absolutely something we see. The work will dictate the workload and requirements of the GPU and that’s how we do processing.

John Wondolowski

For any infrastructure decision, today there are three guiding principles, cyber risk, resilience, and simplification. With some vendors, you can make the choice for a wide variety of reasons based on the use case. “This is better for this, or this will solve that problem.” But in some cases, you may be inventing technical debt two years, three years down the road. Jeff, how do you play into why I need an edge device, but I’ve got to understand the cyber risk and the data protection I’ve got. It’s got to be resilient, and it’s got to help me with my resilience across the enterprise, and I need simplification, I don’t need more devices in other places.

Jeff Guenthner

That’s ultimately what drew me to StorMagic. We had a client that was in the retail space, and I happened to stumble across the use case with Home Depot. I started learning more about what the Home Depot Enterprise IT Operations team was doing with StorMagic, and the fact that they’ve set up their clusters. It’s hardware with StorMagic. So that will take care of my data requirement. They’re using Ansible to deploy. So that takes care of my simplification requirement from an operations perspective, because you’ve got 2200 locations. Then they use an enterprise backup vendor to protect that data at the edge, so it’s immutable. That’s a perfect architecture. I love that. I’m just going to go blueprint the hell out of that because this client and other clients are asking for that in various shapes and sizes. That’s what ultimately led me to come to your doorstep and have those conversations and get engaged.

Bruce Kornfeld

Jeff, you had mentioned to me a few weeks ago that you had a mini case study of a retail customer in your space. Without mentioning their name, do you want to walk through their requirements and how you are able to solve it with our tech?

Jeff Guenthner

This particular client was just in a technical debt situation with age of assets and needing to change over the assets. We were able to do that with some new infrastructure and that they had a VMware license. We use SvSAN with their VMware license. They have a perpetual license, which made things really easy, we just popped SvSAN on there. The cool thing about the project is, they have an amazing project manager, and they were really concerned about scheduling the downtime of the existing infrastructure because of its age and then inserting the new infrastructure that we had created in partnership with you. Doing that in off hours because they just weren’t sure how it was going to go. They did one off hours, went beautifully. Then they just made the decision, well, we’re just going to do this during store business hours. They sent their field technicians out with the new equipment and literally did the changeover during business hours, and nobody noticed a thing. It’s phenomenal. It’s a great thing. They’re in love with your platform. It just works.

Then the other thing for them was simplicity. As John alluded to earlier, when we took a look at the various architectures, because we did consider going to a vSAN, you just reached this point with vSAN where they have a couple hundred locations, you have to create clusters in your vCenter to manage those vSAN deployments. You reach a point where when you get to 64, you have to have another management. You’ve got these managers of managers, and you’ve got somebody managing that and that gets really complex, you have to remember which cluster you got to go to and you have to start thinking about, “well, how am I going to organize my stores?” That’s not simple. It starts to get to be a pain in the butt. But with StorMagic, the thing that the client loved and that we love is from a witness perspective, I can go do a two node and I can have a thousand locations to one witness. That’s just insane scalability, and the premium to do that is not expensive at all. Clients are just shocked, this is all our clients.

Bruce Kornfeld

Thanks for that story. I’ll ask you to think about predictions for the future. Whether you want to think about one year, three year, 10 years. Think about the future technology-wise. Where do you think the industry is going? What do you think customers are worried about for the future?

John Wondolowski

The pay and the pace of technology. The pace of change in technology has always been way too fast, but it’s been accelerating way too fast over the last few years. That’s going to continue. ChatGPT wasn’t even out three years ago, so going out three years from now, I can’t even imagine what that would be. However, for IT leaders the most important thing is getting there without making too many mistakes. There are vendors that have proven to IT leaders that, risk of vendor lock-in is a problem and risk of obsolescence is a problem. I was talking to a bunch of CIOs just two weeks ago and I said, “when you sign a three year contract, what’s your expectation?” They said, “I hate signing three year contracts because you have no idea what’s going to be there three years from now.” I’m not going to make any predictions on the technology, but if you look back and you think of the capabilities that AI has provided on a monthly acceleration scale, anthropic just this week, or last week they say that Claude can code independently for 30 hours. My fear is that, for an enterprise IT leader, the complexity that that kind of progress is going to create in technical debt in shadow IT. We’ll be here three years from now, but there’s going to be a whole host of problems that are going to be difficult to mitigate going in.

Bruce Kornfeld

I want to thank both of you guys for coming. Jeff and John have been great, great conversations talking tech with two CIOs, two CXOs, two CTOs, whatever you call yourselves. We’re all in the tech world.

My name is Bruce Kornfeld, Chief Product Officer at StorMagic, and you’ve been here for PodMagic where we’re talking about real IT challenges and how to solve them. Happy to have you and we’ll see you again next time. Thanks for joining.