>> Live from the MIT campus in Cambridge, Massachusetts, it's the Cube covering the 12th annual MIT Chief Data Officer and Information Quality Symposium. Brought to you by SiliconANGLE Media. >> Welcome back to the Cube's coverage of MIT CDOIQ here in Cambridge, Massachusetts. I'm your host, Rebecca Knight along with my co-host, Peter Burris. We're welcoming back Dr. Santikary who is the Vice President and Chief Data Officer of ERT, thanks for coming back on the program. >> Thank you very much. >> So, in our first interview, we talked about the why and the what and now we're really going to focus on the how. How, what are the kinds of imperatives that ERT needs to build into its platform to accomplish the goals that we talked about earlier? >> Yeah, it's a great question. So, that's where our data and technology pieces come in. As we were talking about, you know, the frustration that the complexity of clinical trials. So, in our platform like we are just drowning in data, because the data is coming from everywhere. They are like real-time data, there is unstructured data, there is binary data such as image data, and they normally don't fit in one data store. They are like different types of data. So, what we have come up with is a unique way to really gather the data real-time in a data lake and we implemented that platform on Amazon Web Services Cloud and that has the ability to ingest as well as integrate data of any volume of any type coming to us at any velocity. So, it's a unique platform and it is already live. Press release came out early part of June and we are very excited about that and it is commercial right now, so yeah. >> But, you're more than just a platform. The product and services on top of that platform, one might say that the services in many respects are what you're really providing to the customers. The services that the platform provides, have I got that right? >> Yes, yes. So, platform like in a uBuild different kinds of services, we call it data products on top of that platform. So, one of the data products is business intelligence where you do real-time decisioning and the product is RBM, Risk Based Monitoring, where you come up with all the risks that a clinical trial may be facing and really expose those risks preemptively. >> So, give us an examples. >> Examples will be like patient visit, for example. A patient may be noncompliant with the protocol, so if that happens, then FDA is not going to like it. So, before they get there, our platform almost warns the sponsors that hey, there is something going on, can you take preemptive actions? Instead of just waiting for the 11th hour and only to find out that you have really missed out on some major things. It's just one example, another could be data quality issues, right? So, let's say there's a gap in data, and/or inconsistent data, or the data is not statistically significant, so you raise some of these with the sponsors so that they can start gathering data that makes sense. Because at the end of the day, data quality is vital for the approval of the drug. If that quality of the data that you are collecting is not good, then what good is the drug? >> So, that also suggests a data governance is gotta be a major feature of some of the services associated with the platform. >> Yes, data governance is key, because that's where you get to know who owns which data, how do you really maintain the quality of data overtime? So, we use both tools, technologies, and processes to really govern the data. And as I was telling you in our session one, that we are the custodian of this data, so we have fiduciary responsibility in some sense to really make sure that the data is ingested properly, gathered properly, integrated properly. And then, we make it available real-time for our real-time decision making, so that our customers can really make the right decisions based on the right information. So, data governance is key. >> One of the things that I believe about medical profession is that it's always been at the vanguard of ethics, social ethics, and increasingly, well, there's always been a correspondence within social ethics and business ethics. I mean ideally, they're very closely aligned. Are you finding that the medical ethics, social medical ethics of privacy and how you handle data, are starting to inform a broader understanding of the issues of privacy, ethical use of data, and how are you guys pushing that envelope if you think that has an important future? >> Yes, that is a great question like we use all these, but we have like data security in place in our platform, right? And the data security in our case plays at multiple level. We don't co-mingle one sponsor's data with others, so they're always like particularized. We partition the data in technical sense and then we have permissions and roles so they will see what they're supposed to be seeing. Not like interdepending on the roles, so yeah, data security is very critical to what we do. We also de-anonymize the data, we don't really store the PII like personally identifiable information as well like e-mail address, or first name or last name, you know? Or social security number for that matter. We don't, when you do analysis, we de-identify the data. >> Are you working with say, European pharmaceuticals as well, Bayer and others? >> Yeah, we have like as I said -- >> So, you have GDPR issues that you have satisfied? >> We have GDPR issues, we have like HIPAA issues, so you name it, so data privacy, data security, data protection, they're all a part of what we do and that's why technology's one piece that we do very well. Another pieces are the compliance, science, because you need all of those three in order to be really, you know, trustworthy to your ultimate customers and in our case they are pharmaceutical companies, medical device companies, and biotechnology companies. >> Where there are lives at stake. >> Exactly. >> So, I know you have worked, Santi, in a number of different industries, I'd love to get your thoughts on what differentiates ERT from your competitors and then, more broadly, what will separate the winners from the losers in this area? >> Yeah, obviously before joining ERT I was the Head of Engineering at Ebay. >> Who? (panel members laughing) >> So, that's the bidding platform, so obviously we were dealing with consumer data, right? So, we were applying artificial intelligence, machine learning, and predictive analytics, all kinds of things to drive the business. In this case, while we are still doing predictive analytics, but the idea of predictive analytics is very different, because in our case here at ERT, we can't recommend anything because they are all like, we can't say hey, don't take Aspirin, take Tylenol, we can't do that, it needs to be driven by doctors. Whereas at Ebay, we would just talking to the end consumers here and we would just predict. >> Again, different ethical considerations. >> Exactly, but in our domain primarily like ERT, ERT is the best of breed in terms of what we do, driving clinical trials and helping our customers and the things that we do best are those three ideas like data collection, obviously the data custodiancy that includes privacy, security, you name it. Another thing we do very well is real-time decisioning that allow our customers, in this case pharmaceutical companies, who will have this integrated dataset in one place, almost like cockpit, where they can see which data is where, what the risks are, how to mitigate those risks, because remember that this trials are happening globally. So, your sites, some sites are here, some sites are in India, who knows where? >> So, the mission control is so critical. >> Critical, time critical. And as well as, you know, cost effective as well, because if you can mitigate those risks before they become problems, you save not only cost, but you shorten the timeline of the study itself. So, your time to market, you know? You reduce that time to market, so that you can go to market faster. >> And you mentioned that it can be as long, the process can be a $3 billion dollar process, so reducing time to market could be a billion dollars a cost and a few billion dollars of revenue, because you get your product out before anybody else. >> Exactly, plus you're helping your end goals which is to help the ultimate patients, right? Because you can bring the drug five years earlier than what you have ended for, then you would save lots of lives there. >> So, the one question I had is we've talked a lot about these various elements, we haven't once mentioned master data management. >> Yes. >> So, give us a little sense of the role that master data management plays within ERT and how you see it changing, because you used to be a very metadata, technical-oriented thing and it's becoming much more something that is almost a reflection of the degree to which an institution has taken up the role that data plays within decision-making and operations. >> Exactly, a great question. At the master data management has people, process, and technology, all three that they co-mingle each other to drive master data management. It's not just about technology. So, in our case, our master data is for example, site, or customers, or vendors, or study, they're master data because they lead in each system. Now, depenation of those entities and semantics of those entities are different in each system. Now, in our platform, when you bring data together from this pair of systems, somehow we need to harmonize these master entities. That's why master data management comes into play. >> While complying with regulatory and ethical requirements. >> Exactly. So, customers for example aren't worried as once said. Or, pick any other name, can be spared 20 different ways in 20 different systems, but when you are bringing the data together, into a called platform, we want nobody to be spared only one way. So that's how you mental the data quality of those master entities. And then obviously we have the technology side of things, we have master data management tools, we have data governance that is allowing data qualities to be established over time. And then that is also allowing us to really help our ultimate customers, who are also seeing the high-quality data set. That's the end goal, whether they can trust the number. And that's the main purpose of our integrated platform that we have just launched on AWS. >> Trust, it's been such a recurring theme in our conversation. The immense trust that the pharmaceutical companies are putting in you, the trust that the patients are putting in the pharmaceutical companies to build and manufacture these drugs. How do you build trust, particularly in this environment? On the main stage they were talking this morning about, how just this very notion of data as an asset. It really requires buy-in, but also trust in that fact. >> Yeah, trust is a two-way street, because it has always been. So, our customers trust us- we trust them. And the way you build the trust is through showing, not through talking, right? So, as I said, in 2017 alone, 60% of the FDA approval went through our platform, so that says something. So customers are seeing the results, they're seeing their drugs are getting approved, we are helping them with compliance, we're artists with science, obviously with tools and technologies. So that's how you build trust, over time, and we have been around since 1977, that helps as well because it says that true and tried methods, we know the procedures, we know the water as they say, and obviously folks like us, we know the modern tools and technologies to expedite the clinical trials. To really gain efficiency within the process itself. >> I'll just add one thing to that, trust- and test you on this- trust is a social asset. At the end of the day it's a social asset. There are a lot of people in the technology industry continuously forget is that they think trust is about your hardware, or it's about something in your infrastructure, or even your applications. You can say you have a trusted asset, but if your customer says you don't, or a partner says you don't, or some group of your employees say you don't, you don't have a trusted asset. Trust is where the technological, the process, and the people really come together, that's the test of whether or not you've really got something the people want. >> Yes, and your results will show that, right. Because at the end of the day, your ultimate test is the results. Everything hinges on that. And the experience helps, as your experience with tools and technologies, signs, regulatories, because it's a multidimensional venn diagram almost, and we are very good at that, and we have been for the past 50 years. >> Well Santi, thank you so much for coming on the program again, it's really fun talking to you. >> Thank you very much, thank you. >> I'm Rebecca Knight for Peter Burris, we will have more from M.I.T CDOIQ in just a little bit.

>> Live from the MIT campus in Cambridge Massachusetts, it's theCube, covering the 12th annual MIT Chief Data Officer and Information Quality Symposium, brought to you by SiliconANGLE media. >> Welcome back to theCUBE's coverage of MIT CDOIQ here in Cambridge, Massachusetts. I'm your host Rebecca Knight along with my co-host Peter Burris. We're welcoming back Dr. Santikary, who is the Vice President and Chief Data Officer of ERT. Thanks for coming back on the program. >> Thank you very much. >> So in our first interview we talked about the why and the what and now we're really going to focus on how, the how. How, what are the kinds of imperatives that ERT needs to build into its platform to accomplish the goals that we talked about earlier. >> Yeah, it's a great question. So, that's where our data and technology pieces come in. We are as we were talking about in our first session that the complexity of clinical trials. So in our platform like we are just drowning in data because the data is coming from everywhere. There are like real-time data, there is unstructured data, there is binary data such as image data and they normally don't fit in one data store. They are like different types of data. So what we have come up with is a unique way to really gather the data real time, in a data lake, and we implemented that platform on Amazon web services ... Cloud and ... that has the ability to ingest as well as integrate data of any volume, of any type coming to us at any velocity. So it's a unique platform and it is already live, press release came out early part of June and we are very excited about that. And it is commercial right now. So, yeah. >> But you're more than just a platform, you're product and services on top of that platform, one might say that the services in many respects are what you're really providing to the customers, the services that the platform provides. Have I got that right? >> Yes, yes. So platform like you build different kinds of services we call it data products on top of that platform. So one of the data products is business intelligence. Why do you do real time decisioning? Another product is RBM, Risk-Based Monitoring, where you ... come up with all the risks that a clinical trial may be facing and really expose those risks preemptively. >> So give us some examples. >> Examples will be like patient visit for example. Patient may be non-compliant with the protocol. So if that happens then FDA is not going to like it. So before they get there our platform almost warns the sponsor that hey there is something going on can you take preemptive actions? Instead of just waiting for the 11th hour and only to find out that you have really missed out on some major things. It's just one example. Another could be data quality issues, right. So let's say there is a gap in data and/or inconsistent data or the data is not statistically significant. So you've to raise some of these with the sponsors so that they can start gathering data that makes sense because at the end of the day, data quality is vital for the approval of the drug. If the quality of the data that you are collecting is not good, then what good is the trial? >> So that also suggested that data governance is got to be a major feature of some of the services associated with the platform. Have I got that right? >> Yes, data governance is key because that's where you get to know who owns which data. How do you really maintain the quality of data over time? So we use both tools, technologies, and processes to really govern the data and as I was telling you in our session one, that we have the custodian of these data. So we have fiduciary responsibility in some sense to really make sure that the data is ingested properly, gathered properly, integrated properly and then we make it available real time for real time decision making so that our customers can really make the right decisions based on the right information. So data governance is key. >> One of the things that I believe about medical profession is that it's always been at the vanguard of ethics, social ethics and increasingly, well there has always been a correspondence between social ethics and business ethics. I mean, ideally they're very closely aligned. Are you finding that the medical ethics, social medical ethics of privacy and how you handle data are starting to inform a broader understanding of the issues of privacy, ethical use of data, and how are you guys pushing that envelope if you think that that is an important feature? >> Yeah, that's a great question. We use all these, but we have like data security in place in our platform, right? And the data security in our case plays at multiple level. We don't co-mingle one sponsor's data with other's. So they are always like particalized. We partition the data in technical sense and then we have permissions and roles. So they will see what they are supposed to be seeing. Not like, you know depending on the roles. So yeah, data security is very critical to what we do. We also de-anonymize the data. We don't really store the PII like Personally Identifiable Information as well like email address or first name or last name or social security number for that matter. When we do analysis, we de-identify the data. >> Are you working with European pharmaceuticals as well, Bayer and others? >> Yeah, we have like as I said. >> So you have GDPR issues (crosstalk). >> We have GDPR issues. We have like HIPPA issues. So you name it. Data privacy, data security, data protection. They are all a part of what we do and that's why technology is one piece that we do very well. Another pieces are the compliance, science. Because you need all of those three in order to be really trustworthy to your ultimate customers and in our case they are pharmaceutical companies, medical device companies, and biotechnology companies. >> Where there are lives at stake. >> Exactly. >> So I know you have worked Santi in a number of different industries. I'd like to get your thoughts on what differentiates ERT from your competitors and then more broadly, what will separate the winners from the losers in this area. >> Yeah, obviously before joining ERT, I was the head of data engineering at eBay. >> Who? (laughing) >> So that's the bidding platform so obviously we were dealing with consumer data right? So we were applying like artificial intelligence, machine learning and predictive analytics. All kinds of thing to drive the business. In this case, while we are still doing predictive analytics but the ideal predictive analytics is very different because in our case here at ERT we can't recommend anything because they are all like we can't say hey don't take Aspirin, take Tylenol. We can't do that. It's to be driven by doctors. Whereas at eBay, we were just talking to the end consumers here and we would just predict. >> Different ethical considerations. >> Exactly. But in our domain primarily like ERT, ERT is the best of breed in terms of what we do, driving clinical trials and helping our customers and the things that we do best are those three areas like data collection. Obviously the data custodiancy that includes privacy, security, you name it. Another thing we do very well is real time decisioning. So that allow our customers, in this case, pharmaceutical companies who will have this integrated dataset in one place. Almost like a cockpit where they can see which data is where, where the risks are, how to mitigate those risks. Because remember that these trials are happening globally. So some sites are here, some sites are in India. Who knows where? >> So the mission control is so critical. >> Critical, time critical. >> Hmm. >> And as well as you know cost-effective as well because if you can mitigate those risks before they become problems, you save not only cost but you shorten the timeline of the study itself. So your time to market, you know. You reduce that time to market so that you can go to market faster. >> And you mentioned that it can be, they could be, the process could be a 3 billion dollar process. So reducing time to market could be a billion dollars of cost and a few billion dollars of revenue because you get your product out before anybody else. >> Exactly. Plus you are helping your end goals which is to help the ultimate patients, right? >> And that too. >> Because if you can bring the drug five years earlier than what- >> Save lives. >> What you had intended for then you know, you'd save lots of lives there. Definitely. >> So the one question I have is we've talked a lot about these various elements. We haven't once mentioned master data management. >> Yes. >> So give us a little sense of the role that master data management plays within ERT and how you see it changing. Because it used to be a very metadata technical oriented thing and it's becoming much more something that is almost a reflection of the degree to which an institution has taken up the role that data plays within decision making and operation. >> Exactly, a great question. The master data management has like people, process, and technology. All three, they co-mingle each other to drive master data management. So it's not just about technology. So in our case, our master data is for example, site or customers, or vendors or study. They're master data because they live in each system. Now definition of those entities and semantics of those entities are different in each system. Now in our platform when you bring data together from disparate systems, somehow we need to harmonize these master entities. That's why master data management- >> While complying with regulatory and ethical requirements. >> Exactly. So customers for example Novartis let's say, or be it any other name, can be spelled 20 different ways in 20 different systems. But when we are bringing the data together into our core platform, we want Novartis to be spelled only one way. So that's how you maintain the data quality of those master entities. And then obviously we have the technology side of things. We have master data management tools. We have data governance that is allowing data qualities to be established over time and then that is also allowing us to really help our ultimate customers who are also seeing the high quality dataset. That's the end goal, whether they can trust the number. And that's the main purpose of our integrated platform that we have just launched on AWS. >> Trust is just, it's been such a recurring theme in our conversation. The immense trust that the pharmaceutical companies are putting in you, the trust that the patients are putting in the pharmaceutical companies to build and manufacture these drugs. How do you build trust, particularly in this environment? We've talked, on the main stage they were talking this morning about how just this very notion of data as an asset, it really requires buy-in, but also trust in that fact. >> Yeah, yeah. Trust is a two-way street, right? Because it has always been. So our customers trust us, we trust them. And the way you build the trust is through showing not through talking, right? So, as I said, in 2017 alone, 60% of the FDA approval went through our platform. So that says something. So customers are seeing the results. So they are seeing their drugs are getting approved. We are helping them with compliance, with audits, with science, obviously with tools and technologies. So that's how you build trust over time. And we have been around since 1977, that helps as well, because it's a ... true and tried method. We know the procedures. We know the water, as they say. And obviously, folks like us, we know the modern tools and technologies to expedite the clinical trials, to really gain efficiency within the process itself. >> I'll just add one thing to that and test you on this. Trust is a social asset. >> Yeah. >> At the end of the day it's a social asset and I think what a lot of people in the technology industry continuously forget, is that they think the trust is about your hardware, or it's about something in your infrastructure, or even in your applications. You can say you have a trusted asset but if your customer says you don't or a partner says you don't or some group of your employees say you don't, you don't have a trusted asset. >> Exactly. >> Trust is where the technological, the process, and the people really come together. >> And the people come together. >> That's the test of whether or not you've really got something that people want. >> Yes. And your results will show that, right? Because at the end of the day, your ultimate test is the results, right? And because that, everything hinges on that. And then the experience helps as you're experienced with tools and technologies, science, regularities. Because it's a multidimensional Venn diagram almost. And we are very good at that and we have been for the past 50 years. >> Great. Well Santi, thank you so much for coming on the program again. >> Okay, thank you very much. >> It was really fun talking to you. >> Thank you. >> I'm Rebecca Knight for Peter Burris. We will have more from MIT CDOIQ in just a little bit. (upbeat futuristic music)

>> Narrator: Live from Washington DC, it's The Cube. Covering .conf2017, brought to you by Splunk. >> And welcome back to .conf2017 here on The Cube. We continue our coverage from the Walter Washington Convention Center. Dave Vellante, John Walls, if you're wondering where we are, I mean physically, the White House is about a mile that way, and the U.S. Capitol is about a mile that way. So we're kind of sandwiched between where it's all happening, Dave. >> Yeah, I mean this exhibit hall is about a mile that way and a mile that way. (laughing) >> Yeah, if you're hungry, leave now for lunch. It's going to be a bit of a hike. We're going to talk about analytics, obviously, at this show, but with Cisco's Altaf Karim, Senior Manager of service line and product lead, so a practice lead. So Altaf, thank you for being with us here. >> You're very welcome. >> Thanks for the time. Let's talk about the Cisco network optimization service, and obviously how that comes into play with analytics, what that's all about. I know that's certainly near and dear to your mission. >> Sure. So as you mentioned, Cisco's network optimization service, it's a consulting-based service offer that we provide to hundreds of customers globally, where we're actually providing some experts in the field of Cisco products. These consultants know Cisco products in and out. Our span reaches globally in many different industries, and what we do is we really work with our customers first, our consultants work with our customers first to identify what sort of business outcomes that they're trying to achieve. These could be related to things like high availability, performance, and then really work from there to understand what types of things need to happen from an assessment standpoint, or architecture, or deployment standpoint, that they can optimize to make the most use of their network. Some of the key benefits of Cisco optimization service are increased productivity for our customers, better user experience, as well as customers who have made an investment in IT. Our consultants are able to work with them and devise a strategy on faster time to value of that investment. So those are some of the key tenets of-- >> Mr. Vellante: So this is a for-pay service, correct? >> Yes. >> Okay, and it starts presumably with an assessment, where you got to get the right people in the room, and maybe you have some automated tooling to help me do discovery, and things like that, and you're maybe looking at machine data and so forth. Take us through the life-cycle of an engagement. Where does it start? How do we engage? How does one engage with you? Where does it start and where does it go? >> Yeah, sure. So, it all starts with our consultants working with our customers first, as I said, to understand what types of business objectives are they trying to accomplish. We then essentially backtrack from there, and understand what things in the network can we control. For example, high availability, process of change management, improved performance on their network, and essentially devise KPIs and metrics that essentially back into the business outcome that they're trying to accomplish. And of course, we have a whole slew of capabilities around analytics, that our consultants bring to the table to essentially become proactive, and help the customer achieve those business outcomes. >> So it might be a customer comes to you and says hey, I'm having problems with my network. It's down too much, it's not performing the way I want. I think it's change management related, you know it probably is, but I don't know where to start. So you bring a tiger team in, and then what? You use all kinds of tooling and other expertise to surface the problem? >> Yeah, sure. So, your question actually delves into what types of KPI can our consultants provide to our customers, to show them how their network is doing, right? And so there's a couple of different ways to do this. One is, you can take a look at what data is available to you, and start to sift through that. And that can be a very cumbersome process that is lengthy. You're really looking for that needle in the haystack to try to figure out what types of insights you can find to make an impact to the business outcome. Another way to approach it is the way we do it from a process standpoint, is inwards from the customer's business outcome. What exactly are we trying to impact? Is it network performance? Is it high availability? And then, our consultants will actually come up with metrics and KPIs based on intellectual capital that our service offer has, and essentially create custom applications based on Splunk, to essentially provide those insights and views and visibility into the network, back to the customer. >> So is it fair to say that Splunk would be the primary ITOM tool, if I can use that term? Splunk doesn't really talk about ITOM, I guess, directly, but to me it's ITOM, IT operations management, but that is the primary platform that you guys would use and deploy? >> I would say that's one of the primary components. Splunk plays a very, very strategic role in how our consultants interact with our customers. So if you think about the premise behind and the value proposition behind network optimization service, is our leading-edge and world-class expertise in networking. And that's what we're known for. And so now when you think about analytics, especially proactive and predictive, you really need the right mixture in ingredients of things to come together, to provide meaningful analytics back to customers. And really, if you think about a trifecta of domain expertise, data science, as well as an understanding of potentially open-source technologies and platforms. But in this case, we're actually strategically using Splunk to play the piece of that last bit. And so what that means is we have consultants who are world-class, leading experts in networking, but we're also training them and asking them to walk a little bit in the shoes of data analysts. And, if you think about an audience or a constituent that is highly technical, quantitative-minded, Splunk is a pretty easy platform for them to learn and start to make an impact by creating custom applications, KPIs, and metrics, for their own customers, that they can use to be proactive and be preemptive, and provide those insights back to the customer. So that's the role that Splunk plays in our service. How much of your business is sort of Aspirin versus vitamin? In other words, how much is it, I got a pain point, I need a tactical solution to that pain point, versus you know what? I'm thinking about re-architecting my network, east west problem, right? Help me think that through, how I sort of transition from my legacy network to a more modernized network. How much is each of those? >> I would say they both play a pretty significant fare. Depending on where the customer is in the life cycle and what they're trying to accomplish, we certainly have a healthy dosage of customers who we work with transactionally, to architect new networks, to deploy new technology, to help them realize their IT spend in a quicker way. But then, a very significant part of our business also is, what do you do on the day two? You can build all this great stuff, right? But if you don't optimize it for peak performance, if you don't optimize it for high availability, or if it's not keeping up with your evolving needs and standards, then you might get in trouble. You're not using the most out of your network. So that's a healthy business as well. >> You mentioned KPIs. What are you tracking? And, what data matters? How do you determine what's relevant, what's not? You know, big problems, or big challenges at least. >> Yeah. That's a very important question, right? And to me, coming from a services background, it's very much rooted in knowing what your domain is about, because as I mentioned before, if you start with all the plethora of data that's available to you, and start to sift through it, you may or may not find something, right? But, our consultants work with the customer and identify what are specific things that we care to monitor, and what are specific KPI that we want to essentially do trending on, or to identify patterns around, so that we can accomplish some sort of business outcome. So for example, if you care about network performance, you're looking at metrics about capacity or bandwidth, or QOS. If you care about customer experience, you're probably, from a wifi standpoint, looking at signal strengths, looking at disassociations, how often and how quickly customers can connect to wifi networks. So really, it depends on what the customer is looking for. And our approach is that we have solid expertise in a number of networking disciplines ranging from routing, switching, wireless, data center, and others. So we have analytic service offers that go deep into each of those technology areas, and we can figure out what KPI to monitor to best achieve that business outcome, but then we also can bring all of that back together and provide that holistic network perspective, and one of the key things that we want to look at, to make sure network is operating optimally. >> Does your practice bleed into the security vector at all? Is that an adjacent area, or is that sort of a main area? >> Yeah. I would say security is paramount for our customers. For the network optimization service, it's actually an adjacent area, but it's definitely something that we work to include into all of our consultative guidance and recommendations to our customers. >> To whom do you sell, I mean, typically? When you initiate an engagement, is it a head of network? Is it a CIO level? And who do you get involved in the sort of initial meeting, and throughout the lifecycle of the project? >> Yeah. That's a really good question, and I would say that it varies depending on what types of analytics that they're also looking for. So let me give you a couple of different examples. So one example is the IT director or IT manager, who is really looking for a tool or analytics, visibility, insights, into how pieces of their network are performing so that they can achieve high availability, increase in network performance, or can better process their change management. So that's one type of buyer. But the other type of buyer is also at the CIO level, which is increasingly also more interested in using analytics to figure out where they are, and benchmark themselves against how others in their industry, or their peers, may be doing. So we've actually started to begun a lot of interesting conversations there, where some of the analytics that we can provide to our customers who opt in, is really rooted around benchmarking how they're doing in different areas such as performance, their software feature, their software or hardware or feature diversity compared to others in their own industry, and really can identify along with our consultative guidance which areas are really important for them to pay attention to, because they're doing something potentially different than everyone else in their industry. >> How about this challenge of IT networks, they're organic, they're constantly changing. So are you coming in, fixing a problem, and then I got to call you back? Or are you teaching me how to fish? >> I would say we're doing a little bit of both. So there's definitely reactive and remediation portions of our service offer. Unfortunately, that happens more than you would like, because you don't think about what to fix until something actually goes wrong. But, one of our flagship service offers, the network optimization services, is all about proactive and optimizing an existing network, so you make sure you're never getting to a place where you end up having to remediate something. And it's not just about remediation or fixing something that's broken, it's really about fine-tuning a well-oiled machine, to make sure that you're getting the most out of your IT investment. >> Yeah, but what kind of a, you talk about machine learning here, capabilities, what do you have in that vein? >> Yeah, so that's a really good question. When we start talking about proactive, and the predictive aspects of our consulting as well as our analytics, machine learning plays a pretty significant role, and I can only expect the contribution that will make to increase exponentially over time. A perfect example, one example of how we use machine learning is actually the machine learning tool kit inside of Splunk. So, if you think about our main premise behind network optimization, is to provide consulting, and provide recommendations on how to optimize the network. But when you think about what a network is, and it's a living and a breathing thing, each network is different, right? No network is the same. So, what machine learning, and especially the machine learning toolkit from Splunk, allows us to do is for a specific customer, it actually allows us to create a baseline of normalcy. What is normal for hundreds and thousands of KPIs and variables, for that specific customer? I think if we asked a human to do that, they'd probably still be going on-- (imitates gunshot) exactly, right? And so, that's an example of how we use machine learning toolkit from Splunk, and not only identifying what is normal for that customer, but then we can use supervised learning to start to identify anomalies and trends and patterns, and really begin to enable our consultants with the data and foresight around what types of things are happening on that network, so that they can in turn be proactive, and be predictive and preemptive in their exchanges with the customer. >> And these services are done on a T&M basis, or a fixed fee, or both? >> They're done both ways. We're pretty flexible, and there's a whole slew of offers outside of what I just talked about, that are available as well. >> What's typical of people? It just depends, right? >> I would say for pinpoint specific things that need to get done, they're more transactional in nature. And then when you're looking for entire lifecycle in a suite of services to help you optimize and be proactive and predictive and preemptive, that's where we have a subscription-based offer that is our optimization offer. >> Okay, and then you guys will actually, well you'll do this mostly remotely, I presume, but you go on site periodically to just impress the flesh and feel-out the culture? >> Absolutely. When we actually start an engagement with a customer, it's quite common for us to go on site, work to get to know the customer, the players, the network, understand what the business outcomes are, make sure that we're devising our deliverables in a way that actually impacts some sort of outcome, and they're not just rooted in some networking measures that don't necessarily make any impact there, right? So that's really important to us. So we definitely go on site. But of course, one of the value propositions of our offer is our intellectual capital. And when we talk about some of the analytics applications that engineers are building for a specific customer, now talk about that happening across hundreds of customers and engineers, devising new ways to create insights and visibilities in their own customer, and the sharing that happens between the engineers, so that they can bring those learning back to their own customer. >> Well, the door's open for business at Cisco, and Altaf Karim, we appreciate your time sharing with us why and how, and what you're doing, and wish you all the best of luck down the road too. Thanks for being with us here, first time on The Cube, right? >> First time on The Cube. >> Alright. >> Thank you for having me. >> You are now an alum. Welcome to the club. >> Great. >> Alright, Altaf Karim, joining us here on The Cube. We'll continue live from Washington D.C., right after this. (electronic theme music)