Among the highlights of the XChange Best of Breed conference were a number of lessons for the AI era from the top CEOs in the business, including a massive infrastructure upgrade opportunity, a shift by more partners into the MSSP market and the continued rise of the marketplace model. Here’s a look at the 10 biggest takeaways from the CEOs of Hewlett Packard Enterprise, Cohesity, CrowdStrike, Cisco Systems, Quantiphi and Ahead.

Solution providers that are willing to step up and make the investments to become AI-proficient will find big opportunities from customers that need help rearchitecting their business to remain competitive in the AI era.

That was among the takeaways from the CEOs at the XChange Best of Breed Conference, hosted by CRN parent The Channel Company, that featured on-stage CRN interviews with Hewlett Packard Enterprise President and CEO Antonio Neri, Cohesity CEO Sanjay Poonen, CrowdStrike CEO George Kurtz and Cisco Systems Chair and CEO Chuck Robbins. The lineup also included on-stage interviews with solution provider executive standouts Asif Hasan, co-founder and president of AI superstar Quantiphi, which has completed 2,500 AI projects with a $2 billion business impact, and Daniel Adamany, founder and CEO of $4 billion solution provider powerhouse Ahead.

“This is going to be a bigger cycle [of opportunity] than the cloud was,” said Neri. “If you’re not on the cycle, you’re going to be suffering down the road because the reality is you won’t have the relevancy that people expect.”

Solution providers need to invest in and work with vendors that make it easy for them to be successful, said Neri. “Work with those vendors who make it easy for you to go sell,” he said. “You can work with so many people. You have to pick the one that is willing to make an investment with you and that you also are willing to make an investment in them. They need to be committed to treat you no different than we do at HPE, which is as an extension of our sales force. We do extensive enablement and also invest in the people to help your brand. But in the end, you have to be very proficient up front on day zero because many enterprise customers want help.”

Neri, in fact, said he just returned from a visit with a customer that had upgraded to a modern network infrastructure but needed help with business use cases. “You can create a lot of opportunity by being that catalyst, including business in IT, being that adviser.”

Furthermore, Neri said many customers do not know how to run the new modern infrastructure and are looking for help.

“Many of these customers don’t know how to run these things on the back end,” he said. “They don’t, especially as you get into these new technologies.”

During Robbins’ interview, he said every solution provider needs to be thinking about building an AI practice with an eye toward understanding AI enterprise uses cases. “Look at where you can add value and the infrastructure that is going to be required [for AI solutions],” he said.

“A massive percentage of our customers don’t feel like their current infrastructure is prepared for the loads and the stress that AI is going to put on their technology infrastructure,” said Robbins.

In fact, he said, customers are diverting money set aside for AI to modernize their infrastructure so they are ready for the new wave of AI applications.

“Stay focused on helping your customers modernize right now, just to be ready,” he said. “But I think today, the best thing you can do is just remain educated. These use cases are going to emerge, and as they do, I think they’re going to move very quickly.”

Robbins recently visited a large energy company in the Middle East where engineers with decades of experience are so in tune with their machines that they can literally listen to a compressor or other equipment and know it will fail within the next 30 days based on the sounds it is emitting.

“So they’re actually taking these sounds, putting them into a training model, and then actually running AI models against the audio that’s coming out of their infrastructure, which is just incredible,” he said. “I think you’re going to see more and more of these things. I think right now the key is staying educated, staying up to speed on these use cases, and understanding and learning the infrastructure stacks.”

Poonen, for his part, urged partners to be aware of “AI washing”—when technology vendors overinflate claims of what their AI technology can do—and to focus on tangible AI use cases as Cohesity itself has done with its Gaia product. Gaia uses retrieval augmented generation (RAG) AI to capture insight from archived secondary data.

“With AI techniques, we are getting insight out of this large amount of data, [literally hundreds of exabytes],” said Poonen.

The most tangible use case, he said, is the ability for a bank or hedge fund to summarize contract terms from the up to 95 percent of secondary data that has been largely unavailable for them to search until now. “What if you could write a query that says, ‘Summarize all my contract terms that I’ve had over all my PDFs and give me a summary of all the contract terms,’” said Poonen.

“Really think through the use cases,” Poonen said. “Don’t AI wash your capabilities. Go look at a use case that is tangible. Test it with customers. It took us a long time to perfect the first use case of Gaia.”

Kurtz, who founded one of the first cybersecurity service companies, said he sees a “huge opportunity” today for solution providers to provide services to secure the entire pipeline around AI.

“There are going to be a lot of solutions that are going to be focused on helping people secure the entire pipeline of AI, from data gathering to data generation to using it,” he said. “I think there’s going to be a lot of consulting work around the fact that you’re going to have to go in with a customer and say, ‘OK, you now want to use AI. How are you going to use it securely? How are you going to buy technologies that are going to help you use it securely and put guardrails around it? I think early on if you have expertise from an AI solution provider perspective, that’s going to grow pretty quick.”

Here are 10 of the most profound takeaways from CRN’s interviews with these industry-leading CEOs.

1. AI Is Sparking A Massive Infrastructure Update

The GenAI revolution is already sparking a massive infrastructure update as customers prepare for a generation of new AI applications to increase productivity and drive revenue gains.

Neri told XChange Best of Breed attendees that the combination of HPE’s $14 billion acquisition of Juniper Networks—the basis for a modern AI networking fabric from edge to cloud—opens up a $135 billion market opportunity.

“This is the first time in the history of both HP and HPE that the company will have the full intellectual property stack, from silicon to the infrastructure to the operating system through the software services and security to provide that modern edge-to-cloud networking fabric, which represents $135 billion more in opportunity,” said Neri.

Every customer is rethinking and retooling their IT architecture for the AI era, said Neri. “There is no question about that, but it will take time,” he said. “We are early. Think about where most of this audience plays in the enterprise. We are early on in that cycle, so there is still a lot of time. But you need to decide who you want to be in that cycle, the sooner the better.”

Robbins said both the need to modernize infrastructure for the AI era and old assets vulnerable in the current threat landscape is causing customers to look for help from partners to upgrade their infrastructure.

“From a cyber perspective, we have so many customers—and I say we as a technology industry, including Cisco—we all know we have customers that sweat their assets and have technology in their infrastructure that’s beyond the last day of support from different vendors that have vulnerabilities that have been announced, with security notices with patches that have been released, and customers haven’t updated them,” he said. “And the adversaries around the world are taking advantage of it, whether it’s critical infrastructure or the carrier networks or whatever.

“I think increasingly, our customers are going to really need us to come in and candidly, help them prioritize, and we’re building some capabilities that we’re going to share with our partners around how do we help our customers prioritize, and look at what to tackle first,” said Robbins. “Because it’s going to become a real issue for our customers if we don’t get this resolved. And I think that as well as the modernization to be ready for AI are two areas that I think are going to drive our customers to continue this upgrade cycle.”

2. GenAI, Next-Gen SIEM Capabilities Pushing More Partners To Become MSSPs

New GenAI capabilities and next-generation SIEM functionality are leading a bevy of new MSSPs to enter the market to help customers beat back the threats from bad actors.

Kurtz told XChange Best of Breed attendees that CrowdStrike’s Charlotte AI product has opened the door to a major Security Operations Center transformation that is reshaping the MSSP landscape.

In fact, Kurtz said Charlotte AI is automating tasks that once took SOC analysts eight hours and reducing them to 10 minutes.

Speaking of the SOC transformation power of Charlotte AI, Kurtz said: “There isn’t a customer that any of these folks in the room work with that doesn’t want to do things better, faster and cheaper, to simplify, to make it easy. People just want easy stuff that works, and they want to reduce costs. AI has the ability to do that.”

Robbins said that Cisco’s $28 billion acquisition of Splunk has brought together SIEM and observability with application performance management to power a massive next-generation SOC opportunity.

“We have billions of pieces of threat intelligence that we’re seeing every day come through all the endpoints and the assets and the network, and we can put all that together in the next-generation SOC, the next-generation operations center, and observability and security are going to come together,” he said. “And so you’re going to have an operations center that is looking at all of these things happening, and you get insight from every aspect of your infrastructure.”

That SOC breakthrough is going to cause many more Cisco partners to become MSSPs, said Robbins.

“That’s absolutely a very real way that customers want to procure these services,” he said. “Increasingly, customers obviously don’t want to do integration work. They want the outcome. They just want to move fast. They want the technology to provide the benefit the technology is supposed to provide. And they’re looking to people like all of you in this room to actually do all the stuff that hasn’t happened to get them to that outcome. … Absolutely they’ll buy it as a service.”

3. The CrowdStrike-Microsoft Détente Aims To Make Windows More Resilient

Kurtz told XChange Best of Breed attendees that the aim of Microsoft and CrowdStrike working more closely together in the wake of the CrowdStrike defective content update on July 19 that led to a “blue screen of death” for 8.5 million Windows users is a more resilient Windows operating system.

When asked specifically if the CrowdStrike and Microsoft partnership would make Windows more “resilient and safer,” Kurtz said: “I think that’s the goal. You know, when you look at the reasons that all security products have to run in the [Windows] kernel, there are four main reasons: No. 1 is detection; No. 2 is prevention, No. 3 is anti-tamper and No. 4r is performance. So if there are other ways to be able to do that with extensions and platform advancements that they have, we’re part of the conversation.”

In fact, Kurtz said he was part of the Sept. 10 Microsoft Windows Endpoint Security Ecosystem Summit at Microsoft’s headquarters in Redmond, Wash., that brought 10 security vendors together in the wake of the CrowdStrike global incident.

The Windows ecosystem “challenge and benefit” is “Microsoft is open, so you have got literally thousands and thousands of drivers and companies that work in this protected mode,” said Kurtz. “So this isn’t just a security discussion. This is how they advance the platform. And we’re a part of providing input, as we have for the last 13 years, working from Windows 7 all the way up to Windows 11.”

As to what Kurtz thinks success will look like for the Microsoft-CrowdStrike partnership over the next two years, Kurtz said: “I think success really focuses on the customer. So is the customer in a better spot? Do they have more resiliency in their architecture? Are there things that Microsoft and its platform can provide with security providers that allow us to do what we need to do? Because you can’t just run in user mode. If you run in user mode, you will just be disabled. So you have to have the kernel element today and the user element, which we have both. But if we and the rest of the community have additional capabilities that the Windows operating system gives us, we’ll always take advantage of that, as we have. So I think just advancing the platform, making it more resilient, and making it have features that we need, but also resilient if there is a failure in any component of the system, whether it’s the application or whether it’s the operating system, just to ensure that the reliability and the resiliency is there for customers.”

4. Driving Data Insight Is Critical To Success In The AI Era

The race is on for partners to help companies derive insight out of the unimaginable amount of data that is stored by businesses in secondary storage backup archives.

Poonen told XChange Best of Breed attendees that Cohesity’s breakthrough Gaia product—which uses RAG AI—is making data useful that up until now has mostly been relegated to tape backups.

The ability to mine that secondary data saves customers time, money and even makes data connections that weren’t possible before, said Poonen.

“No one had done that before,” said Poonen. “No one had. Typically to do this you had to rehydrate the data, take it out of backup—which is highly complex—put it into another file or something like that, and then run an algorithm on that,” he said.

Kurtz, for his part, singled out the ability for his company’s Charlotte AI product to drive breakthrough insight that is making customers more secure.

“At the end of the day, Charlotte AI has the collective wisdom of what CrowdStrike knows over the last 10 years, all our intel reports, etc, plus it has the ability to actually do work on behalf of our customers,” he said.

5. A Disturbing New Threat Vector: Bad Actors Being Hired As IT Pros

Kurtz told XChange Best of Breed attendees that his company worked with law enforcement to uncover a North Korean government effort to put in place bad actors working as IT pros in more than 100 large corporations.

“We actually discovered a lot of it, worked with law enforcement well before anything was public,” he said. “We actually found a concerted movement by the People’s Republic of North Korea that was actually focused on having technology folks get jobs into 100-plus different huge companies out there, and we were able to pick this up with our OverWatch service, [which provides 24×7 AI-powered threat hunting intelligence across all domains].”

One troubling aspect of the North Korean operation, said Kurtz, was employees were hired remotely, which resulted in laptop farms with KVMs attached that were controlled from North Korea.

“We just saw something that seemed weird in one company,” said Kurtz. “The power of CrowdStrike is we can look across the entire customer base, and we’re like, ‘This seems really weird,’ and then we start digging into it. And it turns out that these operatives were being hired by companies, and then essentially laptop farms were set up.”

The CrowdStrike team effectively uncovered that “Bob Smith in Texas isn’t really Bob Smith,” said Kurtz. “It’s a laptop with KVMs attached to it so they can be controlled remotely from North Korea. So we actually had to go to customers and say, ‘Hey, we think you have a [bad actor] insider [in your company], which is never an easy thing to do,” he said.

“We caught this like100 times,” he said. “It’s a testament to our technology, the Overwatch team and the threat and [intelligence-]focused approach that we brought to market and pioneered.”

6. A Networking Battle For The Ages With Big Architectural Differences

Both HPE with its impending acquisition of Juniper Networks and Cisco with its acquisition of Splunk are offering partners and customers two divergent paths in the AI networking era.

Neri, for his part, told XChange Best of Breed attendees that the Juniper acquisition effectively remakes HPE as a “networking company at its core,” something he said that “Cisco has forgotten for a little bit.”

Neri said the Juniper Networks acquisition—which brings with it Juniper’s Mist AI technology—is aimed at building a modern, AI-driven networking experience. “Obviously Splunk plays more in the SIEM space,” he said. “That’s not what we do, and we would never do that. Our focus is to really bring together a series of amazing assets.

Among the HPE-Juniper architectural integration points, Neri said, will be the merging of Juniper’s “amazing QFX Fabric” with the HPE Slingshot architecture, which has been used to scale tens of thousands of GPUs in HPE supercomputers. “We believe with our high-performance silicon we can reduce the complexity of the platform,” he said.

Once the Juniper deal closes, expected late this year or early 2025, HPE will have an $11 billion networking business, representing more than one-third of HPE’s sales and probably 50 percent of its profits, said Neri.

Robbins, for his part, said Cisco is providing the networking, security and observability that customers are demanding for the AI era.

“You look at the combination of networking and security and the importance of those two coming together—which [HPE does] not have—and you look at data center infrastructure, you look at campus networking with wireless, with all of the observability, the security and everything that we have—I mean, we have more technology that brings more value to our customers in the infrastructure layer than anybody else,” said Robbins.

“We are the only company who can bring networking, observability, security to our customers, all integrated together,” said Robbins. “And we think that all of those things coming together are more important than they ever have been, and that’s what we’re going to deliver. We are a secure networking company that actually delivers incredible capabilities, whether it’s an AI-ready data center, future-proof workplace, or an underlying layer of digital resilience that we’re going to deliver.”

Cisco is poised to bring its Cisco Nexus HyperFabric, which is slated to be available early next year, to the enterprise with Cisco networking, Vast storage, Nvidia GPUs and Cisco CPUs with cloud-based orchestration and life-cycle management all integrated with security. “We want to really help our customers more easily deploy these AI applications as these use cases become more apparent,” he said.

As for Neri’s claim that Cisco has “forgotten” its core networking mission, Robbins said simply: “Well, we forgot more about networking than they’ll probably know about networking.”

7. Scale, Speed And Innovation Are Key To Success In AI Era

Cohesity’s Poonen told attendees that his company’s acquisition of Veritas combines the “scale” of Veritas— which takes Cohesity from $546 million to $2 billion in revenue—with the “speed and innovation” of Cohesity. “Think of them like a BMW and us like a Tesla,” said Poonen.

The deal, which is slated to be completed by the end of the year, doubles Cohesity’s research and development team and packs a big global punch, with approximately 45 percent of Veritas’ business outside the U.S.

“If we can take the scale that we will have as an almost $2 billion company and still move fast,” said Poonen, then it opens the door to create an “iconic” company that could grow rapidly from $2 billion to $5 billion.

Every dollar of Cohesity revenue opens up as much as two times that in consulting and professional services along with product licensing and reselling opportunities for partners, said Poonen.

As for cloud service provider opportunities, Poonen said, there is a huge “AI-powered data security market opportunity.”

8. AI Solutions Require Heavy Up-Front Investments By Solution Providers

Quantiphi’s Hasan told XChange Best of Breed attendees that the AI model requires higher up-front R&D investments but also pays off in higher margins.

In the past several years, Quantiphi’s margins have increased by about 300 basis points because of the company’s software and services AI focus, with a lower cost of service delivery than partners with an infrastructure-focused channel model, said Hasan.

Approximately 95 percent of Quantiphi’s sales come from software and services, with 60 percent of that from building AI applications and 40 percent from assisting customers in modernizing their data and applications estate for AI.

“What that means is when you compare us with a traditional channel business, we are an asset-light company, which means that we don’t carry any inventory because we’re not doing much reselling,” he said. “But at the same time, we also have a lot of intangible assets: our software, our portfolio, patents.”

Quantiphi has higher up-front R&D costs because of the big investments it makes in its AI partnerships with Nvidia, Google Cloud and Amazon Web Services, said Hasan. But the cost of goods sold in terms of “how much human capital is going into the service delivery” is less because of the heavy emphasis on software rather than infrastructure delivery, he said.

One example of the AI benefits being brought to customers by Quantiphi is as much of a 60 percent decrease in help desk calls by employing AI agents that are able to answer questions for customers, said Hasan.

Quantiphi is also seeing dramatic gains in software engineering productivity through the use of AI agents and big boosts in personal productivity with content summarization, said Hasan.

9. Ability To Successfully Integrate Acquisitions Is A Big Differentiator In AI Era

Adamany of Ahead, the $4 billion solution provider powerhouse that has made 10 acquisitions since 2019, told XChange Best of Breed attendees that successful integration of acquisitions requires a collaborative, value-driven approach rather than a “roll-up” strategy mindset.

“You can really [mess] up an acquisition,” said Adamany. “The biggest way you can [mess] it up is people. Because in our business, we’re a services business. It’s all about the people. So, you can flush money down the toilet pretty quickly if you buy a company and you don’t approach it right, you don’t integrate it right. If you lose all the people, you’re worse off than when you started.”

Adamany said he sees the process as learning from the company being acquired and making sure that those new employees feel like they are “valued” and “happy” to be a part of the combined company.

“I often say that as we are doing an acquisition, my goal is that in a year from now you will be glad that we acquired your company,” he said. “If I can get 90 percent of the people to feel that way, then they’re going to stay. They’re going to be excited about the future. If they are [unhappy] that we bought them, and they miss the old days and they don’t like the way it is they’re eventually going to leave, or they are going to be a cancer within the organization.”

Adamany said companies that feel they absolutely must do an acquisition to continue to grow are setting themselves up for failure.

“That’s a really bad position to be in because it’s a position of weakness,” he said. “I feel like those acquisitions don’t typically go well because you’re basically marrying somebody to make you feel better. You’ve got to be happy yourself before you take that on.”

10. Marketplace Ecosystem Momentum Is Accelerating With Innovative New Programs

The marketplace ecosystem model continues to build momentum with new twists, including a CrowdStrike model aimed at getting customers to adopt more modules in the 28-module Falcon platform without having to go back to procurement for approval of a new purchase order.

The FalconFlex model backed by a CrowdStrike financial services subsidiary was “born” out of a demand by customers for the security vendor to make it easier to adopt the platform without having to go back to procurement for each module. The FalconFlex model also offers higher discounts in a cloud-credit-like model.

“It’s a commitment and burn down model, very similar to a cloud provider like AWS,” said Kurtz. “Essentially a customer will commit any number of dollars. In some cases, you are talking about millions of dollars in commitment. We’ve done over $700 million of deal value with Falcon Flex already, and it is less than a year old.”

The commitment opens up the ability for customers to quickly adopt any one of the Falcon modules, said Kurtz.

The CrowdStrike FalconFlex model comes on top of highly successful CrowdStrike partnerships with AWS and Google Cloud marketplaces.

In fact, the AWS marketplace model has resulted in more than $1 billion in sales for CrowdStrike.

“This marketplace model is an area that’s going to continue to drive growth for folks in the room,” he said. “Customers want it. It reduces a lot of friction. You can have enterprise contracts that you can check the box on and get up and running really quickly. We’ve done deals on some of these marketplaces in less than an hour. So for the partners here, working with these marketplaces is being part of what customers want. They want to go in that direction. And a lot of times they have credits they have to burn down. So you want to be a part of it, not on the outside looking in at it.”