Extreme President and CEO Ed Meyercord and Nabil Bukhari, chief product and technology officer, speak with CRN about how AI and GenAI are shaping Extreme’s strategy, recent networking M&A activity — including the Cisco-Splunk acquisition and the pending HPE-Juniper deal — and how Extreme’s ‘One Network, One Cloud’ approach is helping the specialist stand out.

Extreme Networks, a networking player that may be smaller from a revenue perspective compared with some of the networking market heavyweights, can’t be counted out.

The company that’s proven to be a mighty rival thanks to its loyal base of channel partners, enterprise and large-venue customers, including many of the major NFL and MLB stadiums, was once again dubbed a Leader in the most recent Gartner Magic Quadrant for Enterprise Wired and Wireless LAN Infrastructure report among Cisco Systems, Hewlett Packard Enterprise and Juniper Networks. Speaking of the competition, with the networking industry in flux as evidenced by Cisco’s $28 billion acquisition of Splunk and the pending deal between HPE and Juniper Networks, Extreme is making hay of what it calls the market “confusion” by doubling down on its winning strategy and not getting distracted, the companies top leaders told CRN ahead of its annual Extreme Connect user conference this week.

Extreme took to the event to highlight some impressive modern connectivity use cases, including one customer that’s bringing connectivity to remote areas such as the Amazon. The company also unveiled Extreme Labs, a new hub for research, development and innovation in networking.

Extreme President and CEO Ed Meyercord and Nabil Bukhari, Extreme’s chief product and technology officer, spoke with CRN about the new value chain that AI presents to networking vendors and channel partners and laid out how AI is shaping Extreme’s strategy this year. The executives also spoke candidly about what the recent networking M&A activity—namely, the Cisco-Splunk acquisition and the pending HPE-Juniper deal—means for Extreme, and how Extreme’s “One Network, One Cloud” approach is helping the specialist stand out.

What follows are excerpts from the conversation.

How are networking trends like AI shaping Extreme’s strategy?

Meyercord: This is a really exciting time in networking. Networking is effectively a cloud service today. Security, with the enterprise becoming distributed [because of] COVID and then [with] more devices on the network and the challenges of a cloud-driven network, is moving rapidly to the cloud. We have purpose-built AI tools in our network and now generative AI is on the scene and it’s creating all kinds of new opportunities for us. These are the areas [in which] we’re investing, and we’re moving very fast. The major themes are [with] this convergence of networking and security and generative AI tools, how can we work with all these amazing customers doing amazing things in various geo[graphies] and verticals to help them accelerate?

We’re building the future together. We have a very healthy mix of partners, and all of our initiatives are together with our partners. As we look at the solutions that we’re building and we look at the technology, particularly as it relates to AI, we [now] have Extreme Labs. Here again, this is us working together with partners and new technology as it relates to AI, and our AI experts, that’s going to be in the lab. This is a collaboration with Extreme and partners and ecosystem partners where we’re open and we’re sharing [and] we’re investing. It’s very much a joint effort with us and our partners.

Bukhari: The relationship between the vendors and partners is going to drastically change with technologies like AI. If you produce an AP [access point] and a switch and you give it to the partner to rack and stack it, that was a value chain that has existed in networking for a long period of time, but with technologies like AI, that’s going to have to get reformed. The way to think about this is that as vendors, our core strength is innovation and technology and building that core platform. The partners are the ones that are closest to the customer. So, when you put those two things together, that is when you really deliver that value of AI [with] customized, specific use cases for customers. That’s what Extreme Labs is. Extreme Labs is our internal innovation center where we are working on technologies that are really in the hype cycle, like AI. Late last year, we decided that we [were] doing this innovation ourselves, but we want to pull in our channel partners as well as our ecosystem partners. So, we created this program called Extreme Labs. Now, we’re announcing it publicly. But underneath that umbrella, this is what we do. We are putting in our ecosystem partners like Microsoft, Intel … and we’re building this core technology and including them in there. But then we have also our channel partners, and the logic there is that not only are we providing products like Extreme AI Expert that the users can do directly, we’re [also] introducing an entire AI studio and AI core platforms that our partners can then build AI services on top. And that is just absolutely unique. There’s nobody else that is doing it and we’re investing a lot of money in doing that. We currently have a big innovation challenge going on with our channel partners and they’re co-developing with us. To me, the value chain of the future when it comes to AI is there are big [large language model] providers, then there are vendors like us that will build an AI platform on it and then enable our partners to deliver those super precise AI applications to the end customers. That’s the value chain. And I think in that direction, we are farther along than anybody else.

What does the recent M&A activity — the Cisco-Splunk acquisition and the pending HPE-Juniper deal — mean for Extreme?

Bukhari: While there’s a lot of progress in terms of technology, there’s a lot of confusion when you look at the vendor space in networking. The top one is a little busy with its massive acquisition of Splunk, and then No. 2 and No. 3, while brilliant companies, they have a long road ahead of them in trying to figure out what they want to be, which essentially translates to risk. It’s risk for not only the customers, but risk for partners. They might not be in that space where they were previously before all these changes in the vendor landscape. We are right now the company in networking. Our vision is clear, our execution is clear. We’re not distracted with some acquisition or something else. I feel that’s translating into a lot of interest from the broader partner space as well. We are the only ones with the least risk. That’s the feedback that we’re getting from the broader partner base. And it’s a unique opportunity for us.

Meyercord: We know a thing or two about M&A because we’ve done so many acquisitions. Cisco closing Splunk; they’re less than 50 percent networking today. Their eye is moving further and further away from that core business where we’re focused. And we’re hearing that from the channel and we’re hearing that from customers. And then an M&A deal of the size of what’s being done between HPE and Juniper, especially considering the massive price that was paid, means a huge round of costs have to come out and there is going to be a lot of synergy that comes out of that, so they have to make lots of cuts. They also have to decide on whose partner program survives. There are economic ramifications to that. What’s the technology portfolio? What’s going to survive? And how are we going to drive this? Engineers are going to be fighting over this for years and they’ll disagree. So, at the end of the day, in terms of our position and our focus, we’re getting a lot of good feedback from customers and partners and we’re pretty excited about the next couple years and just kind of the dynamic of the competitive landscape.

What makes Extreme’s One Cloud platform approach stand out in the market?

Meyercord: We’ve been working on building a platform to allow our partners to deliver managed services, effectively making it very easy for them to leverage our services with a unique approach in terms of licensing as far as consumption billing, which doesn’t exist in the industry. We’re the only player that has a true consumption model for licenses. But there’s a lot of interest from partners in this. And as we look at the evolution of services, as they evolve and become part of effectively, new services and new licenses, I think this is probably a platform that’s going to be in high demand. Because this is what people want to use.

Bukhari: [Some] of our MSP partners, they have already built their services on the networking side. And we are giving them limited availability for our universal ZTNA and our MSP partners are picking it up right away and including that into their products and solutions to their market. And the reason why they’re doing this is very simple. One is, everything that we do, whether that is networking and security that we are kind of blending and converging, and AI, they’re all under the same cloud and the same commercial model. That is absolutely unique. There is nobody else that has their entire portfolio under one cloud, one commercial model. And when you are an MSP, that is what you need. You need lower transaction costs, you will need lower operational costs so that you can scale. We introduced this a year ago and we’ve been working with some core partners to do that. But now we’re really getting into the stride of it. And we’re starting to see that scale up. This really stems from our vision of that One Network, One Cloud. It’s one network that connects people, application devices, no matter where they are. That vision is translating into that success.

How is Extreme approaching AI and how does that help the company differentiate in the networking space?

Bukhari: AI has been around for a long period of time, and we all built our AI products on it—[Juniper Networks] Mist built Marvis on it, we built our CoPilot on it, and those are all AI products. But what is new is generative AI, which kind of changes the playing field because it behaves differently and learns a little bit differently. The scale is just unprecedented. When it comes to generative AI, I think we are kind of in that round of everybody announcing their products, so, we plan on differentiating in the following three ways.

First and foremost, we view generative AI as an ecosystem play because if you have data from one source, you’re valuable. If you have data from two sources, you’re 100 times more valuable. And if your data is from three sources, you’re 10,000 times more valuable. It’s exponential. So, we’re not going out with the idea like some of the other people have that a generative AI is a product. We’re going with generative AI is an ecosystem, it’s a platform and the more data sets, which means more partners that are plugged into it, the more valuable it becomes. And we’re spending a lot of money in making that easier and more deployed.

The second part would be the trust factor. How would the customer actually trust it? Underneath that is data privacy, data residency, controls, accuracy, all of those things that are playing in. We all know that when you’re just looking at an LLM, the accuracy is somewhere like 55 [percent] to 56 percent. Well, that’s OK when I’m asking ChatGPT for a recipe to fry eggs. But when I’m building my network on it, that’s just not acceptable. So, building in that layer of controls [and] protections around it, and building it in such a way that not only your data is protected, but the answers that you get are verified with expert information and expert models.

The third [area] where we are differentiating is that quite frankly, generative AI allows us to go way beyond AIOps. The applications of generative AI are not just in terms of managing the network and then telling you where anomalies are. We are already doing that. So, our vision is that with this ecosystem, we are able to allow use cases that go way beyond AIOps. We categorize it into three categories. One is your knowledge questions. So, just simply the fact that you can ask how to deploy Wi-Fi, how to align your licenses, and so forth. But not just from public information. This includes public information, internal knowledge bases, best practices, direct information from our engineering into the models that are constantly getting better. The second set of capabilities, or intelligence or operational queries, is where we can give you and we can provide you information based on real-time data. Real-time data from networks, from security, from storage and so on, and that’s really where ours starts to shine. The last one is going to be “what if” scenario planning. Let’s just say I’m a retailer, and I want to know, ‘How will this Black Friday go for my network and my capacity?’ Previously, you had to build a lot of rule-based models. With our AI models, we can do all of those predictions well in advance. We can give you how to actually configure or change your network for that and we can actually do that for you as well. So, it’s really based on knowledge, intelligence and scenario-predictive use cases. And this goes way, way beyond AI ops. As our ecosystem of partners grows, we are able to add multiple other data sets. That’s our vision. Our vision is that AI is a platform ecosystem play. AI will live or die with trust, and that AI must go way beyond AIOps. These are the three parts of our strategy that are a little bit unique, and we will absolutely differentiate on each one of them.