(triumphant music) >> Hello and welcome to a special CUBE Conversation, exclusive content here in Palo Alto Studios, I'm John Furrier, the co-founder of SiliconANGLE Media, and cohost of theCUBE. We have exclusive news with Vectra Networks announcing new funding, new R and D facility. I'm here with the president and CEO, Hitesh Sheth, who's the president and CEO. Welcome to theCUBE Conversation, congratulations. >> Thank you John. glad to be here. >> So you've got some big news. >> Vectra Networks, you guys doing some pretty cool stuff with AI and cyber. >> Correct. >> But it's not just software, it's really kind of changing the game with IT operations, the entire Cloud movement, DevOps automations, all impacting the enterprise. >> Hitesh: Yes. >> And other companies. >> Hitesh: Yes. >> Before we dig into some of the exclusive news you guys have, take a minute to talk about, what is Vectra? What is Vectra Networks? >> Maybe it'd be useful to give you context of the way we see the security industry evolving. And if you think about the last 20 years, and if you were to speak to the security person in an enterprise, their primary concern would be around access banishment, who gets in, who gets out. The firewall industry was born to solve this problem. And you know, in many ways its been a gift that's kept on giving. You know, you've got companies with multi-billion dollar evaluations, Palo Alto, Checkpoint, Fortinet, you know, piece of Cisco, etc, right? There's roughly about 40 billion dollars on the market cap sitting in this industry today. Now, if you go back to the same enterprise today, and you look at the next 5-10 years and you ask them, "What is the number one issue that you care about?" Right? It's no longer who's getting in and out from an access policy standpoint, it's all about threat, management, and mitigation. So, the threat's signal is now the most important commodity inside the enterprise and the pervasive challenge for the customer, the enterprise customer, is, "How do I get my hands on this threat's signal in the most efficient way possible?" And we, at Vectra, are all about automating and helping our customers hunt for advanced cyber attacks using artificial intelligence. >> Where did you get the idea of AI's automation? I've always said in theCUBE, "Oh, AI's a bunch of b.s. Because real true AI is there. But again, AI is really kind of growing out of machine learning. >> Hitesh: Right >> Automating, and so this kind of loose definition but certainly is very sexy right now. People love AI. >> Hitesh: Correct. >> I mean, AI is awesome. But at a practical matter, it seems to be very important for good things, also for the enterprise, where'd you get the idea for using AI for cyber? >> Well, you know, I would go back to in my journey intersection with the notion of using AI for cyber security, Back in about 2010, there are major cyber events reported in the press. At that time, I was in the networking sector and in the networking sector, we all looked at it and said, "You know, we can do something about this," and being good networking company is, we thought we would build chips that would do DPI and do packet inspection. It was, too be blunt, old school thinking, okay? Fast forward to 2012 and I was sitting with Vinod Khosla of Khosla ventures and we were talking about the notion of security. How can you transform security dramatically >> Mhmm. >> Hitesh: And this is when we started talking about using artificial intelligence. It was very nascent and frankly, if you went up and down Sand Hill at that time, you know, most of the venture companies would have- and they did, because we were raising money at the time, they would look at us and said, "You guys are nuts. This is just not going to happen." You know, it's very experimental, it would take forever to come to pass. But that's usually the best time to go and build a new business and take a risk, right? And we said, you know what, AI has matured enough. >> By the way, at that time, they were also poo-pooing the Cloud. >> Absolutely. >> Amazon will be nothing. >> Yeah, exactly. Generally, a good time, a good time to go and do something revolutionary. But, here are the other things to know. Not only had the technology around AI and its applicability had advanced enough, but two other things have happened at the same time. The cost of compute had changed dramatically. The cost of storage had changed dramatically. And ultimately, if AI is going to be efficient, not only is the software got to be good, but the computer's got to be valid as well. Storage got to be valid as well. These three things were really coming together on their timeframe. >> Well, what's interesting, let's dig into that for a second because knowing what the scene was with networking at the time, you said, "old thinking," but the state of the art, you know, In the 90's and 2000's was, hardware got advanced, so you had wire speed capability. So, you can do some cool things like, you know, like still move through the network and do some inspection. >> Hitesh: Correct. >> And you said DPACK is recommended But that's the concept of looking at the data. >> Hitesh: That's correct. >> John: So, okay, now they might have been narrow view so now you take it back >> Hitesh: Yes. >> With AI, am I getting it right? You're thinking of zooming out saying, okay, >> Hitesh: A couple of things. >> You find that notion of inspection of data >> Right. >> With more storage, more compute >> But it comes down to also, you know, what data are you looking at, right? When you had wire spec in booties, you would apply your classic signature based approaches. So you could deal with known attacks, right? What is really happening, like 2011-2012 onwards is, the attack landscape is more stored dramatically. It changes so fast that the approach of just dealing with the known was never going to be enough. >> Yeah. >> So, how do you deal with the unknown? You need software that can learn. You need software that can adapt on the fly. And this is where machine learning comes into play. >> You got to assume everyone's a bad actor at that point. >> You got to assume everybody has been infiltrated in some way or fashion. >> Well, the Cloud, certainly, you guys were on the front end, kind of probably thought we're crazy with other VC's, you mentioned that. But at the time, I do remember when Cloud was kind of looked at as just nonsense. >> Yeah >> But if you then go look at what that impact has been, you're in the right side of history, congratulations,. What really happened? When was the C change? You mentioned 2012, was that because of the overall threat landscape change? Was that because of open source? Was that because of new state sponsored threats? >> Hitesh: Yeah. A couple things. >> What was the key flash point? >> Hitesh: A couple of things. We saw, at the time, that there was an emerging class of threats in the marketplace being sponsored by either state actors but we also saw that there was significant funding going into creating organized entities that were going to go and hack large enterprises. >> John: Not state sponsored directly, state sponsored, kind of, you know, >> On the side. >> Yeah, on the side. >> Let's call them, "For Profit Entities," okay? >> Sounds like Equifax to me. (laughter) >> That's a good point. And we saw that happening. Trend two was, there were enough public on the record, hacks are getting reported, right? Sony would be a really good example at the time. But just as fundamentally, it's not just enough that there's a market. The technology has got to be sufficiently ready to be transformative, and this is the whole point around what we saw in compute and storage and the fact that there was enough advancement in the machine learning itself that it was worth taking a risk and experimenting to see what's going to happen. And in our journey, I can tell you, it took us about 18 months, really, to kind of tune what we were doing because we tried and we failed for 18 months before we kind of came to an answer that was actually going to gel and work for the customers. >> And what's interesting is having a pattern oriented to look for the unknown >> Hitesh: Yeah >> Because it's, you know, in the old days was, "Hey, here's a bunch of threats, look for'em and be prepared to deploy." Here, you got to deal with a couple of the unknown potentially attack. But also I would say that we've observed the surface areas increased. So, you mention Checkpoint in these firewalls. >> Hitesh: Yes. Absolutely. >> Those are perimeter based security models. So you got a perimeter based environment. >> Hitesh: Correct. >> Everyday. >> Hitesh: And you got IOT. >> IOT. So it's a hacker's dream. >> It's absolutely. The way I like to think about it is you got an end by end probatational issue. You got an infinite possible, if you're a hacker, you're absolutely right, it's Nirvana. You've got endless opportunities to break into the enterprise today. It's just going to get better. It's absolutely going to get better for them. >> John: Well, let's get to the hard news. You guys have an announcement. You've got new funding >> Hitesh: Yeah. >> And an R and D facility, in your words, what is the announcement? Share the data. >> We're really excited to announced that we have raised closed a round of 36 million dollars, Series D funding, it's being led by Atlantic Bridge, they are a growth fund, and they've got significant European roots, and in addition to Atlantic Bridge, we're bringing on board two new investors, two additional investors. The Ireland's Strategic Investment Fund, number one, effectively the sovereign fund of Ireland, and then secondly, Nissho Electronics of Japan. This is going to bring our double funding to 123 millions dollars, today. What we're going to be using this funds for is to find things with. One is the classic expansion of sales and marketing. I think we've had very significance success in our business. From 2016 to 2017, our business grew 181% year end year, subscription based, all subscription revenue. So, we're going to use this, this new fuel, to drive business growth, but just as important, we're going to drive our needs growth significantly. And as part of this new funding, we are opening up a brand new R & D center in Dublin, Ireland. This is our fourth R & D center. We've got one here in San Jose, California. We've got one in Austin, Texas, Cambridge, Massachusetts, and so this is number four. >> John: So, you hired some really smart people. How many engineers do you guys have? >> So, we are about a 140% company, roughly half the company is in R and D. >> I see a lot of engineering going on and you need it, too. So let's talk about competitors. Darktrace is out there, heavily funded companies, >> Hitesh: Yes. >> Their competitor, how do you compare against the competition and why do you think you'll be winning? >> I can tell you, statistically, whether it is Darktrace or we run into barcoding with Cisco as well. We win into large enterprise. We win 90% of the time. [Overlapping Conversation] >> It's actually correct. And I'll describe to you why is it that we win. We look at people like Darktrace and there are other smaller players in the marketplace as well And I'll tell you one thing fundamentally true about the competitive landscape and that differentiates us. AI is on everybody's lips nowadays, right? As you pointed out. But what is generally true for most companies doing AI and I think this is true for our competition as well, it tends to be human augmented AI. It's not really AI, right? This is sort of like the Wizard of Oz, you know, somebody behind the curtain actually doing the work and that ultimately does not deliver the promise of AI and automation to the customer. The one thing we have been very - >> John: They're using AI to cover up essentially manual business models for all people added, is that what you're saying? >> Hitesh: That's correct. Effectively, it's still people oriented answer for the customer and if AI is really true, then automation has got to be the forefront and if automation is really going to be true, then the user experience of the software has got to be second to none >> John: So, I know Mike Lynch is on the board of that company, Darktrace, he was indicted or charged with fraud to front for HP for billions of dollars. So, is he involved? Is he a figurehead? How does he relate to that? >> I think you should talk to Mike. You should put him in this chair and have this conversation. I recommend it, that would be great. >> John: I don't think he'd come on. >> But my understanding is that he has a very heavy hand in the reign of Darktrace. Darktrace, if you go to their website, so this is all public data, if you look at their management chain, this is all Autonomy people. What that means, respect to how Autonomy was running and how Vectra is being run, is for them to speak about, what I can tell you is that, when we meet them competitively, we meet other competitors. >> John: I mean, if I'm a customer, I would have a lot of fear and certainty in doubt to work with an Autonomy led because they had such a head fake with the HP deal and how they handled that software and just software stack wasn't that great either. So, I mean, I would be concerned about that. [Overlapping Discussion] >> History may be repeating itself. >> Okay, so you won't answer the question. Okay, well, let's get back to Vectra. Some interesting, notable things I discovered was, you guys had been observing what's been reported in the press with the Olympics. >> Hitesh: Correct. >> You have information and insight on what's going on with the Olympics. Apparently, they were hacked. Obviously, it's in Korea, so it's Asia, there's no DNS that doesn't have certificates that have been hacked or whatever so, I mean, what's going on in South Korea with the Olympics? What's the impact? What's the data? >> Hitesh: Well, I'm going to think, what is really remarkable is that, despite the history of different kinds of attacks, Equifax, what have you, nation state events, political elections getting impacted and so forth, once again, a very public event. We have had a massive breach and they've been able to infiltrate their systems and the remarkable thing is they- >> John: There's proof on this? >> There's proof on this. This is in the press. There's no secret data in our part, which is, this very much out there, in the public arena, they have been sitting in the infrastructure of the Olympics, in Korea, for months and the remarkable thing is, why were they able to get in? Well, I can tell you, I'm pretty sure that the approach to security that these people took is no different than the approach of security most enterprises take. Right? The thing that should really concern us all is that they chose to attack, they chose to infiltrate, but they actually paused before really fundamentally damaging the infrastructure. It goes to show you that they are demonstrating control. I can come in. I can do what I want for as long as I want. I can stop when I want. >> John: They were undetected. >> They were undetected. Absolutely. >> John: And they realized that these attacks reflected that. >> Absolutely. And given the fact there seems to be a recent trend of going after public events, we have many other such public events coming to bear. >> How would you guys have helped? >> The way we would help them, most fundamentally is that, look, here's the fundamental reality, there are, as we've discussed just a second ago, there are infinite options as to break in, into the infrastructure, but once you're in, right? For people like you and I, who are networking people, you're on our turf and the things you can do inside the network are actually very visible. They're very visible, right? It's like somebody breaking through your door, once they get in, their footprints are everywhere, right? And if you had the ability to get your hands on those footprints, right? You can actually contain the attack at- as close to real time as possible, before any real damage is done. >> But then we're going to see where the action is, no doubt about it, you can actually roll that data up and that's where the computer- >> And then you could apply machine learning. You can extract the data, look at the network, extract the right data out of it, apply machine learning or AI and you can get your hands on the attack well before it does any real damage. >> John: And so to your point, if I get this right, if I hear ya properly, computers are much stronger now. >> Hitesh: Correct. >> And with software and AI techniques, you can move on this data quickly. >> Hitesh: Correct. But you have got to, you've got to have a fundamental mindset shift, which is, "I'm not in the business of stopping attacks anymore, I should try, but I recognize I will be breached every single time. So, then, I better have the mechanisms and the means to catch the attack once it's in my environment." And that mindset shift is not pervasive. I am 1,000% sure at the Olympics that people designed the security search have said, "We can stop this stuff, don't worry about it." You had that taught differently that would not be in this position today. >> This is the problem. In all society, whether it's a shooting at a school or Olympic hack event, the role of data is super critical. That's the focus, thanks for coming on and sharing the exclusive news at theCUBE with exclusive coverage of the breaking news of the new round of funding for Vectra Networks. I'm John Furrier. Thanks for watching. >> Hitesh: Thank you, John. (triumphant music)