 From the SiliconANGLE Media Office in Boston, Massachusetts, it's theCUBE. Now, here's your host, Stu Miniman. Hi, I'm Stu Miniman, and this is CUBE Conversation from our Boston Area studio. Happy to welcome to the program a first-time guest on the program, but from NetScout, who we've been digging into the concept of visibility without borders. Dr. Vikram Saxena, who's with the Office of the CTO from the aforementioned NetScout, thank you so much for joining us. Thanks, Stu, thanks for having me. All right, Dr. Saxena, before we get into kind of your role, why don't you go back, give us a little bit about your background. You and I have some shared background from COM. We both work for some of the arms of Ma Bell back in the day. You were a little bit more senior, and probably a lot more patents than I have, which my current count is still there, so yeah. Sure, happy to do that. You're right, I started in 82, which was two years before the breakup of Ma Bell, so, you know, and then everything started happening right around that time. So yeah, I started in Bell Labs, you know, stayed there close to 20 years, did a lot of the early pioneering work on packet switching before the days of internet, frame relay, all of that happened. It was a pretty exciting time. I was there building up, we built up the AT&T business from scratch to a billion dollars in the IP space, you know, in a voice company that was always challenging. So, and then I moved on to do startups in the broadband space, did two of them, moved to the Boston area, and then moved on to play the CTO role in public companies, saw on a SNETWORKS TEL Labs, and then, you know, came to Net Scout about five years ago. Yeah, you know, I love talking about, you know, some of those incubators of innovation. Bell Labs, you know, historically speaking, just, you know, threw off so much technology. That's right. Been seeing so much in the media lately about, you know, the 50th anniversary of Apollo 11. That's right. And so many things that came out of NASA, Bell Labs was one of those places that helped inspire me to study engineering. That's right. You know, definitely got me on my career, but here we are, 2019. That's right. You're still, you know, working with some of these telcos and how they're all, you know, dealing with this wave of cloud and, you know, the constant change there. So, bring us inside, you know, what's your role inside Net Scout that offers the CTO? Yes. So, Net Scout is in the business of, you know, mining network data. And what we excel at is extracting what we call actionable intelligence from network traffic, which we use to term smart data. But essentially, my role is really to be the bridge between our technology group and the customers. You know, bring out, understand the problems, the challenges that our customers are facing, and then work with the teams to build the right product to, you know, to fit into the current environment. Okay. One of our favorite things on theCUBE is, you know, talking to customers that are going through their transformation. That's right. If you talk about the enterprise, you know, digital transformation, we think there's more than just the buzzword there. Yeah, I've talked to financial institutions, manufacturing, you know, you name it out there. If it's a company that's not necessarily born in the cloud, they are undergoing that digital transformation. Bring us inside, you know, your customer base, that this telcos, the service providers, you know, most of them have a heavy tech component to what they're doing, but you know, are they embracing digital transformation? What does it mean for them? So, you know, as you said, it's a big term that catches a lot of things. But in one word, if I describe for the telcos, it's all about agility. If you look at the telco model historically, it has been on a path where services get rolled out every six months, year, multiple years, you know, not exactly what we call an agile environment compared to today. You know, but when the cloud happened, it changed the landscape because cloud not only created a new way of delivering services, but also changed expectations on how fast things can happen. And that created high expectations on the customer side, which in turn started putting pressure on the telcos and the service providers to become as agile as cloud providers. And as you know, the network, which is really the main asset of a service provider, was built around platforms that were not really designed to be programmable. You know, so they came in with hardwired services and they would change at a very low time scale. And building around that is the whole software layer of OSSBSS, which over time became very monolithic, very slow to change. So coupling the network and the software layer created a very slow moving environment. So this is what's really causing the change to go to a model where the networks can be programmable, which essentially means moving from a hardware-centric model to a software-centric model where services can be programmed on demand and created on the fly and maybe sometimes even under control of the customers. And layering on top of that, changing the OSSS infrastructure to make it more predictive, make it more actionable and driven by advances in machine learning and artificial intelligence to make this entire environment extremely dynamic and agile. So that's kind of what we are seeing in the marketplace. Yeah, I totally agree. That agility is usually the first thing put forward. I need to be faster. Used to be, you know, faster, better, cheaper. Now, like faster, faster, faster, can actually help compensate for some of those other pieces there. Of course, service providers usually, you know, very conscious on the cost of things there because if they can lower their cost, they can usually make them more competitive and pass that along to their ultimate consumers. You know, bring us inside the, you know, you mentioned the change to software that's going on. You know, there are so many waves of change going on there. Everything from, you know, you talk about IoT and edge computing. Yeah. The big, you know, massive roll out of 5G that, you know, even gets talked about in the general press these days and at government states there. So, you know, where are, you know, your customers today? What are some of the critical challenge they have? And, you know, where is that kind of monitoring observability that kind of piece fit in? So, good. So let me give two backdrop points. First of all, you mentioned cost. So they're always very cost conscious, trying to drive it down. And the reason for that is the traditional services have been heavily commoditized. You know, voice, texting, video data, they've been commoditized. So the customers want the same stuff cheaper and cheaper and cheaper all the time, right? So that puts a pressure on margins and reducing cost. But now the industry is at a point where I think the telcos need to grow the top line. You know, that's a challenge because you can always reduce cost, but at some point you get to a point of diminishing returns. So now I think the challenge is how do they grow their top line? You know, so they can become healthier again in that context. And that leads to whole notion of what services they need to innovate on. So it's all about once you have a programmable network and a software that is intelligent and smart, that becomes a platform for delivering new services. So this is where, you know, you see on the enterprise side, SD-WAN, enterprise IoT, all these services are coming now using technologies of software-defined networking, network function virtualization. And 5G, as you mentioned, is the next generation of wireless technology that is coming on board right now, and that opens up the possibility for the first time. Two new things, dimensions come into play. First, not only a consumer-centric focus, which was always there, but now opening it up to enterprises and businesses and IoT. And secondly, fixed broadband, right? The era where telcos used to either drive copper or fiber, slow, cumbersome, takes a lot of time, right? And the cable guys have already done that with coaxial cable. So they need to go faster and faster means use wireless. And finally, with 5G, you have a technology that can deliver fixed broadband, which means all the high-definition video, voice data and other services like AR, VR, into the home. So it's opening up a new possibility rather than having a separate fixed network and a separate wireless network. For the first time, they can collapse that into one common platform and go after both fixed and mobile, and both consumers and enterprise verticals. Yeah, it was one of the big topics of conversation at Cisco Live. It was out in San Diego just a short time ago. It was 5G and then it's Wi-Fi 6, the next generation of that, because I'm still going to need inside my building for the companies, but the 5G holds the promise to give me just so much faster bandwidth, so much dense for environment. I guess some of the concerns I hear out there, and maybe you can tell me kind of where we are and where the telcos fit in is, 5G from a technology standpoint, we understand where it is, but that rollout is going to take time. And it's great to say you're going to have this dense and highly available thing, but that's going to start the same place all the previous generations are. It's the place where actually we don't have bad connectivity today. It's in the urban areas. It's where we have dense populations. Sometimes it's thrown out there, oh, 5G is going to be great for Edge and IoT, and it's like, well, we don't have balloons and planes and the towers everywhere. So where are we with that rollout of 5G? What side of time frames are your customer base looking as to where that goes to play? So I think from what I'm seeing in the marketplace, I think there is a less of a focus on building out ubiquitous coverage, because when the focus is on consumers, you need coverage because they are everywhere, right? But I think where they are focusing on because they want to create new revenue and new top line growth, they are focusing more on industry verticals, IoT. Now that allows you to build out networks and pockets of where your customers are because enterprises are always focused in the top cities and top metro areas. So before you make it available for consumers, if you get an opportunity to build out, at least in the major metropolitan area infrastructure, where you're getting paid as you're building it out, because you're signing up this enterprise customers who are willing to pay for these IoT services, you get paid, you get to build out the infrastructure, and then slowly as new applications emerge, I think you can make it widely available for consumers. I think the challenge on consumer side is the smartphones have been tapped out, and people are not going to get that excited about 5G just to use the next gen iPhone, right? So there it has to be about new applications and services and things that people talk about always on the horizon are AR, VR and things like that, but they're out there, they're not there today, because a device has to come on board that becomes mass consumable and exciting to customers. So while the industry is waiting for that to happen, I think there's a great opportunity right now to turn up services for enterprise verticals in the IoT space, because the devices are ready, and everybody, because enterprises are going through their own digital transformation, they want to be in a connected world, right? So they're putting pressure on telcos to connect all their devices into the network, and there is a monetization opportunity there. So I think what the carriers are going to do is sign up verticals, whether it's transportation, healthcare, so if they sign up a bunch of hospitals, they're going to deploy infrastructure in that area to sign up hospitals. If they're going to sign up manufacturing, they're going to build their infrastructure in those areas where they're there, right? So by that model, you can build out a 5G network that is concentrated on their customer base and then get to ubiquitous coverage later when the consumer applications come. Yeah, so I like that a lot, because when I think back, if we've learned from the sins of the past, it used to be, if we build it, they will come. Dig trenches across all the highways and lay as much fiber as we can, and then the dot-com burst happens, and we have all of this capacity that we can't give away. What it sounds like you're describing is really a service-centric view. I've got customers and I've got applications and I'm going to build to that and then I can build off of that piece there. Can you talk a little bit about that focus and where your customers are going? Yeah, so maybe just slightly before that, what I want to talk about the distributed nature of the 5G network. So you mentioned edge, right? So one of the things that are happening when you want to deliver low latency services or high bandwidth services, you need to push things closer to the edge. As you know, when cloud started, it's more in what we call the core, the large data centers, the hyper-scale data centers where applications are being deployed now. But when you demand low latency, let's say sub-15 millisecond, 10 millisecond latency, that has to be pushed much more closer to the customer. Now, this is what's forcing the edge cloud deployment in 5G. And then what that does is, it also forces you to distribute functionality. Everything is not centralized in the core, but it's distributed in the edge and the core. The control plane may be in the core, but the user plane moves to the edge. So that changes the entire flow of traffic and services in a 5G network. They are no longer centralized. It means it becomes more challenging to be able to manage and assure these services in a highly distributed telco cloud environment which has this notion of edge and core. Now, on top of that, if you say that, this is all about top-line growth and customer satisfaction, then your focus on operationalizing these services has to change from a network-centric view to a service-centric view. Because in the past, as you know, when we were both in Bell Labs and AT&T, we were pretty much focused on the network, focused on the data from the network, the network elements, the switches and the routers and all of that and making sure that the network is healthy. Now, that is good, but it's not sufficient to guarantee that the services and the service level agreements for customers are being met. So what you need to do is focus at the service layer much more so than you were doing it in the past. So that changes the paradigm on what data you need to use, how you want to use it and how do you stitch together this view in a highly distributed environment and do it in real time and do it all very quickly so the customers don't see the pain if anything breaks and actually be more proactive, in a lot of cases, be more predictive and take corrective actions before they impact services. So this is the challenge and clearly from a net scout point of view, I think we are right in the center of this hurricane and given the history, we sort of have figured out on how to do this. Yeah, the networking has a long history of we've got a lot of data, we've got all of these flows and things change, but right exactly as you said, understanding what's happened to that application. That is what we can really tie to make sure it's not just IT sitting on the side but IT driving that business, those applications, those data flows. So yeah, maybe expand a little bit more, net scouts fit there and why it's so critical for what customers need today. Yeah, happy to do that. So if you look at what are the sources of data that you actually can use and what you should use. So basically they fall into three buckets. What I call, first is what I call infrastructure data, which is all about data you get from hypervisors, V-switches, they're telling you more about how the infrastructure is behaving where you need to add more horsepower, CPUs, memory, storage and so on. So that is very infrastructure centric. The second one is from network elements. You know what the DNS servers give you, DHCP servers, what your routers and switches are giving you, the firewalls are giving you and they're also in a way telling you more about what the network elements are seeing. So there's a little bit of a hybrid between infrastructure and a service layer component but the problem with that data is it's very vendor dependent. It's highly fragmented across because they're no real standards on how to create this data. So there is telemetry data, there is syslogs and they all vendors do it what they think is best for them. So the challenge then becomes on the service provider side and how do you stitch together? Because service is an end-to-end construct or an application, it starts at a user and goes to a server and you need to be able to get that holistic view end-to-end. So the most appropriate data that NetScout feels is what we call the wire data or the traffic data is actually looking at packets themselves because they give you the most direct knowledge about how the service is behaving, how it's performing and not only that you can actually predict problems as opposed to react to problems because you can trend this data, you can apply machine learning to this data and be able to say what might go wrong and be able to take corrective action. So we feel that extracting the right contextual information, relevant information, timely information in a vendor independent way, in a way that is universally available from edge to core those are the attributes of wire data and we excel in processing that at the source in real time and converting all of that into actionable intelligence that is very analytics and automation friendly. So this is our strength and what that allows us to do is as they are going through this transition between 4G and 5G between physical and virtual across fixed and mobile networks you can go through this transition if you have a stitch together end to end view that crosses these boundaries or borders as we call it visibility without borders and in this context your operations people never lose insight into what's going on with their customer applications and behavior. So they can go through this migration with confidence that they will not negatively impact their user experience by using our technology. You know, we've thrown out these terms intelligence and automation for decades in our industry but if you look at these hybrid environments and all of these change and come out if an operator doesn't have tools like this they can't keep up. They can't keep up. So I need to have that machine learning I have to have those tools that can help me intelligently attack these pieces otherwise there's no way I can do it. Yeah and one point there is you know it's like garbage and garbage out if you don't get the right data you can have the most sophisticated machine learning but it's not going to predict the right answer. So the quality of data is very important just as the quality of your analytics and your algorithms. So we feel that the combination of right data and the right analytics is how you're going to get advantage of accurate predictions and automation around the whole suite. Okay love that. Right data, right information, right solution. Why don't want to give you right analytics. I want to give you the final word final takeaways for your customers today. So I think we are in a very exciting time in the industry. You know 5G as a technology is probably the first generation technology which is coming on board where there is so much focus on things like security and new applications and so on. And I think it's an exciting time for service providers to take advantage of this platform and then be able to use it to deliver new services and ultimately see their top lines grow which we all want in the industry because if they are successful then we as suppliers, you know, do well, you know. So I think it's a pretty exciting time and we as NetScout are happy to be in this spot right now and to see and help our customers go through this transition. All right, Dr. Vikram Saksena, thank you so much for joining us, sharing with us everything that's happening in your space and glad to see the excitement still with the journey that you've been on. Thank you Stu, happy to be here. All right, and as always check out thecube.net for all of our content. I'm Stu Miniman and thanks as always for watching theCUBE.