- [Announcer] Live, from New York, it's theCUBE. Covering theCUBE New York City 2018. Brought to you by SiliconANGLE Media, and its ecosystem partners. - Hello everyone, welcome back to CUBE NYC. This is a CUBE special presentation of something that we've done now for the past couple of years. IBM has sponsored an influencer panel on some of the hottest topics in the industry, and of course, there's no hotter topic right now than AI. So, we've got nine of the top influencers in the AI space, and we're in Hell's Kitchen, and it's going to get hot in here. (laughing) And these guys, we're going to cover the gamut. So, first of all, folks, thanks so much for joining us today, really, as John said earlier, we love the collaboration with you all, and we'll definitely see you on social after the fact. I'm Dave Vellante, with my cohost for this session, Peter Burris, and again, thank you to IBM for sponsoring this and organizing this. IBM has a big event down here, in conjunction with Strata, called Change the Game, Winning with AI. We run theCUBE NYC, we've been here all week. So, here's the format. I'm going to kick it off, and then we'll see where it goes. So, I'm going to introduce each of the panelists, and then ask you guys to answer a question, I'm sorry, first, tell us a little bit about yourself, briefly, and then answer one of the following questions. Two big themes that have come up this week. One has been, because this is our ninth year covering what used to be Hadoop World, which kind of morphed into big data. Question is, AI, big data, same wine, new bottle? Or is it really substantive, and driving business value? So, that's one question to ponder. The other one is, you've heard the term, the phrase, data is the new oil. Is data really the new oil? Wonder what you think about that? Okay, so, Chris Penn, let's start with you. Chris is cofounder of Trust Insight, long time CUBE alum, and friend. Thanks for coming on. Tell us a little bit about yourself, and then pick one of those questions. - Sure, we're a data science consulting firm. We're an IBM business partner. When it comes to "data is the new oil," I love that expression because it's completely accurate. Crude oil is useless, you have to extract it out of the ground, refine it, and then bring it to distribution. Data is the same way, where you have to have developers and data architects get the data out. You need data scientists and tools, like Watson Studio, to refine it, and then you need to put it into production, and that's where marketing technologists, technologists, business analytics folks, and tools like Watson Machine Learning help bring the data and make it useful. - Okay, great, thank you. Tony Flath is a tech and media consultant, focus on cloud and cyber security, welcome. - Thank you. - Tell us a little bit about yourself and your thoughts on one of those questions. - Sure thing, well, thanks so much for having us on this show, really appreciate it. My background is in cloud, cyber security, and certainly in emerging tech with artificial intelligence. Certainly touched it from a cyber security play, how you can use machine learning, machine control, for better controlling security across the gamut. But I'll touch on your question about wine, is it a new bottle, new wine? Where does this come from, from artificial intelligence? And I really see it as a whole new wine that is coming along. When you look at emerging technology, and you look at all the deep learning that's happening, it's going just beyond being able to machine learn and know what's happening, it's making some meaning to that data. And things are being done with that data, from robotics, from automation, from all kinds of different things, where we're at a point in society where data, our technology is getting beyond us. Prior to this, it's always been command and control. You control data from a keyboard. Well, this is passing us. So, my passion and perspective on this is, the humanization of it, of IT. How do you ensure that people are in that process, right? - Excellent, and we're going to come back and talk about that. - Thanks so much. - Carla Gentry, @DataNerd? Great to see you live, as opposed to just in the ether on Twitter. Data scientist, and owner of Analytical Solution. Welcome, your thoughts? - Thank you for having us. Mine is, is data the new oil? And I'd like to rephrase that is, data equals human lives. So, with all the other artificial intelligence and everything that's going on, and all the algorithms and models that's being created, we have to think about things being biased, being fair, and understand that this data has impacts on people's lives. - Great. Steve Ardire, my paisan. - Paisan. - AI startup adviser, welcome, thanks for coming to theCUBE. - Thanks Dave. So, uh, my first career was geology, and I view AI as the new oil, but data is the new oil, but AI is the refinery. I've used that many times before. In fact, really, I've moved from just AI to augmented intelligence. So, augmented intelligence is really the way forward. This was a presentation I gave at IBM Think last spring, has almost 100,000 impressions right now, and the fundamental reason why is machines can attend to vastly more information than humans, but you still need humans in the loop, and we can talk about what they're bringing in terms of common sense reasoning, because big data does the who, what, when, and where, but not the why, and why is really the Holy Grail for causal analysis and reasoning. - Excellent, Bob Hayes, Business Over Broadway, welcome, great to see you again. - Thanks for having me. So, my background is in psychology, industrial psychology, and I'm interested in things like customer experience, data science, machine learning, so forth. And I'll answer the question around big data versus AI. And I think there's other terms we could talk about, big data, data science, machine learning, AI. And to me, it's kind of all the same. It's always been about analytics, and getting value from your data, big, small, what have you. And there's subtle differences among those terms. Machine learning is just about making a prediction, and knowing if things are classified correctly. Data science is more about understanding why things work, and understanding maybe the ethics behind it, what variables are predicting that outcome. But still, it's all the same thing, it's all about using data in a way that we can get value from that, as a society, in residences. - Excellent, thank you. Theo Lau, founder of Unconventional Ventures. What's your story? - Yeah, so, my background is driving technology innovation. So, together with my partner, what our work does is we work with organizations to try to help them leverage technology to drive systematic financial wellness. We connect founders, startup founders, with funders, we help them get money in the ecosystem. We also work with them to look at, how do we leverage emerging technology to do something good for the society. So, very much on point to what Bob was saying about. So when I look at AI, it is not new, right, it's been around for quite a while. But what's different is the amount of technological power that we have allow us to do so much more than what we were able to do before. And so, what my mantra is, great ideas can come from anywhere in the society, but it's our job to be able to leverage technology to shine a spotlight on people who can use this to do something different, to help seniors in our country to do better in their financial planning. - Okay, so, in your mind, it's not just a same wine, new bottle, it's more substantive than that. - [Theo] It's more substantive, it's a much better bottle. - Karen Lopez, senior project manager for Architect InfoAdvisors, welcome. - Thank you. So, I'm DataChick on twitter, and so that kind of tells my focus is that I'm here, I also call myself a data evangelist, and that means I'm there at organizations helping stand up for the data, because to me, that's the proxy for standing up for the people, and the places and the events that that data describes. That means I have a focus on security, data privacy and protection as well. And I'm going to kind of combine your two questions about whether data is the new wine bottle, I think is the combination. Oh, see, now I'm talking about alcohol. (laughing) But anyway, you know, all analogies are imperfect, so whether we say it's the new wine, or, you know, same wine, or whether it's oil, is that the analogy's good for both of them, but unlike oil, the amount of data's just growing like crazy, and the oil, we know at some point, I kind of doubt that we're going to hit peak data where we have not enough data, like we're going to do with oil. But that says to me that, how did we get here with big data, with machine learning and AI? And from my point of view, as someone who's been focused on data for 35 years, we have hit this perfect storm of open source technologies, cloud architectures and cloud services, data innovation, that if we didn't have those, we wouldn't be talking about large machine learning and deep learning-type things. So, because we have all these things coming together at the same time, we're now at explosions of data, which means we also have to protect them, and protect the people from doing harm with data, we need to do data for good things, and all of that. - Great, definite differences, we're not running out of data, data's like the terrible tribbles. (laughing) - Yes, but it's very cuddly, data is. - Yeah, cuddly data. Mark Lynd, founder of Relevant Track? - That's right. - I like the name. What's your story? - Well, thank you, and it actually plays into what my interest is. It's mainly around AI in enterprise operations and cyber security. You know, these teams that are in enterprise operations both, it can be sales, marketing, all the way through the organization, as well as cyber security, they're often under-sourced. And they need, what Steve pointed out, they need augmented intelligence, they need to take AI, the big data, all the information they have, and make use of that in a way where they're able to, even though they're under-sourced, make some use and some value for the organization, you know, make better use of the resources they have to grow and support the strategic goals of the organization. And oftentimes, when you get to budgeting, it doesn't really align, you know, you're short people, you're short time, but the data continues to grow, as Karen pointed out. So, when you take those together, using AI to augment, provided augmented intelligence, to help them get through that data, make real tangible decisions based on information versus just raw data, especially around cyber security, which is a big hit right now, is really a great place to be, and there's a lot of stuff going on, and a lot of exciting stuff in that area. - Great, thank you. Kevin L. Jackson, author and founder of GovCloud. GovCloud, that's big. - Yeah, GovCloud Network. Thank you very much for having me on the show. Up and working on cloud computing, initially in the federal government, with the intelligence community, as they adopted cloud computing for a lot of the nation's major missions. And what has happened is now I'm working a lot with commercial organizations and with the security of that data. And I'm going to sort of, on your questions, piggyback on Karen. There was a time when you would get a couple of bottles of wine, and they would come in, and you would savor that wine, and sip it, and it would take a few days to get through it, and you would enjoy it. The problem now is that you don't get a couple of bottles of wine into your house, you get two or three tankers of data. So, it's not that it's a new wine, you're just getting a lot of it. And the infrastructures that you need, before you could have a couple of computers, and a couple of people, now you need cloud, you need automated infrastructures, you need huge capabilities, and artificial intelligence and AI, it's what we can use as the tool on top of these huge infrastructures to drink that, you know. - Fire hose of wine. - Fire hose of wine. (laughs) - Everybody's having a good time. - Everybody's having a great time. (laughs) - Yeah, things are booming right now. Excellent, well, thank you all for those intros. Peter, I want to ask you a question. So, I heard there's some similarities and some definite differences with regard to data being the new oil. You have a perspective on this, and I wonder if you could inject it into the conversation. - Sure, so, the perspective that we take in a lot of conversations, a lot of folks here in theCUBE, what we've learned, and I'll kind of answer both questions a little bit. First off, on the question of data as the new oil, we definitely think that data is the new asset that business is going to be built on, in fact, our perspective is that there really is a difference between business and digital business, and that difference is data as an asset. And if you want to understand data transformation, you understand the degree to which businesses reinstitutionalizing work, reorganizing its people, reestablishing its mission around what you can do with data as an asset. The difference between data and oil is that oil still follows the economics of scarcity. Data is one of those things, you can copy it, you can share it, you can easily corrupt it, you can mess it up, you can do all kinds of awful things with it if you're not careful. And it's that core fundamental proposition that as an asset, when we think about cyber security, we think, in many respects, that is the approach to how we can go about privatizing data so that we can predict who's actually going to be able to appropriate returns on it. So, it's a good analogy, but as you said, it's not entirely perfect, but it's not perfect in a really fundamental way. It's not following the laws of scarcity, and that has an enormous effect. - In other words, I could put oil in my car, or I could put oil in my house, but I can't put the same oil in both. - Can't put it in both places. And now, the issue of the wine, I think it's, we think that it is, in fact, it is a new wine, and very simple abstraction, or generalization we come up with is the issue of agency. That analytics has historically not taken on agency, it hasn't acted on behalf of the brand. AI is going to act on behalf of the brand. Now, you're going to need both of them, you can't separate them. - A lot of implications there in terms of bias. - Absolutely. - In terms of privacy. You have a thought, here, Chris? - Well, the scarcity is our compute power, and our ability for us to process it. I mean, it's the same as oil, there's a ton of oil under the ground, right, we can't get to it as efficiently, or without severe environmental consequences to use it. Yeah, when you use it, it's transformed, but our scarcity is compute power, and our ability to use it intelligently. - Or even when you find it. I have data, I can apply it to six different applications, I have oil, I can apply it to one, and that's going to matter in how we think about work. - But one thing I'd like to add, sort of, you're talking about data as an asset. The issue we're having right now is we're trying to learn how to manage that asset. Artificial intelligence is a way of managing that asset, and that's important if you're going to use and leverage big data. - Yeah, but see, everybody's talking about the quantity, the quantity, it's not always the quantity. You know, we can have just oodles and oodles of data, but if it's not clean data, if it's not alphanumeric data, which is what's needed for machine learning. So, having lots of data is great, but you have to think about the signal versus the noise. So, sometimes you get so much data, you're looking at over-fitting, sometimes you get so much data, you're looking at biases within the data. So, it's not the amount of data, it's the, now that we have all of this data, making sure that we look at relevant data, to make sure we look at clean data. - One more thought, and we have a lot to cover, I want to get inside your big brain. - I was just thinking about it from a cyber security perspective, one of my customers, they were looking at the data that just comes from the perimeter, your firewalls, routers, all of that, and then not even looking internally, just the perimeter alone, and the amount of data being pulled off of those. And then trying to correlate that data so it makes some type of business sense, or they can determine if there's incidents that may happen, and take a predictive action, or threats that might be there because they haven't taken a certain action prior, it's overwhelming to them. So, having AI now, to be able to go through the logs to look at, and there's so many different types of data that come to those logs, but being able to pull that information, as well as looking at end points, and all that, and people's houses, which are an extension of the network oftentimes, it's an amazing amount of data, and they're only looking at a small portion today because they know, there's not enough resources, there's not enough trained people to do all that work. So, AI is doing a wonderful way of doing that. And some of the tools now are starting to mature and be sophisticated enough where they provide that augmented intelligence that Steve talked about earlier. - So, it's complicated. There's infrastructure, there's security, there's a lot of software, there's skills, and on and on. At IBM Think this year, Ginni Rometty talked about, there were a couple of themes, one was augmented intelligence, that was something that was clear. She also talked a lot about privacy, and you own your data, etc. One of the things that struck me was her discussion about incumbent disruptors. So, if you look at the top five companies, roughly, Facebook with fake news has dropped down a little bit, but top five companies in terms of market cap in the US. They're data companies, all right. Apple just hit a trillion, Amazon, Google, etc. How do those incumbents close the gap? Is that concept of incumbent disruptors actually something that is being put into practice? I mean, you guys work with a lot of practitioners. How are they going to close that gap with the data haves, meaning data at their core of their business, versus the data have-nots, it's not that they don't have a lot of data, but it's in silos, it's hard to get to? - Yeah, I got one more thing, so, you know, these companies, and whoever's going to be big next is, you have a digital persona, whether you want it or not. So, if you live in a farm out in the middle of Oklahoma, you still have a digital persona, people are collecting data on you, they're putting profiles of you, and the big companies know about you, and people that first interact with you, they're going to know that you have this digital persona. Personal AI, when AI from these companies could be used simply and easily, from a personal deal, to fill in those gaps, and to have a digital persona that supports your family, your growth, both personal and professional growth, and those type of things, there's a lot of applications for AI on a personal, enterprise, even small business, that have not been done yet, but the data is being collected now. So, you talk about the oil, the oil is being built right now, lots, and lots, and lots of it. It's the applications to use that, and turn that into something personally, professionally, educationally, powerful, that's what's missing. But it's coming. - Thank you, so, I'll add to that, and in answer to your question you raised. So, one example we always used in banking is, if you look at the big banks, right, and then you look at from a consumer perspective, and there's a lot of talk about Amazon being a bank. But the thing is, Amazon doesn't need to be a bank, they provide banking services, from a consumer perspective they don't really care if you're a bank or you're not a bank, but what's different between Amazon and some of the banks is that Amazon, like you say, has a lot of data, and they know how to make use of the data to offer something as relevant that consumers want. Whereas banks, they have a lot of data, but they're all silos, right. So, it's not just a matter of whether or not you have the data, it's also, can you actually access it and make something useful out of it so that you can create something that consumers want? Because otherwise, you're just a pipe. - Totally agree, like, when you look at it from a perspective of, there's a lot of terms out there, digital transformation is thrown out so much, right, and go to cloud, and you migrate to cloud, and you're going to take everything over, but really, when you look at it, and you both touched on it, it's the economics. You have to look at the data from an economics perspective, and how do you make some kind of way to take this data meaningful to your customers, that's going to work effectively for them, that they're going to drive? So, when you look at the big, big cloud providers, I think the push in things that's going to happen in the next few years is there's just going to be a bigger migration to public cloud. So then, between those, they have to differentiate themselves. Obvious is artificial intelligence, in a way that makes it easy to aggregate data from across platforms, to aggregate data from multi-cloud, effectively. To use that data in a meaningful way that's going to drive, not only better decisions for your business, and better outcomes, but drives our opportunities for customers, drives opportunities for employees and how they work. We're at a really interesting point in technology where we get to tell technology what to do. It's going beyond us, it's no longer what we're telling it to do, it's going to go beyond us. So, how we effectively manage that is going to be where we see that data flow, and those big five or big four, really take that to the next level. - Now, one of the things that Ginni Rometty said was, I forget the exact step, but it was like, 80% of the data, is not searchable. Kind of implying that it's sitting somewhere behind a firewall, presumably on somebody's premises. So, it was kind of interesting. You're talking about, certainly, a lot of momentum for public cloud, but at the same time, a lot of data is going to stay where it is. - Yeah, we're assuming that a lot of this data is just sitting there, available and ready, and we look at the desperate, or disparate kind of database situation, where you have 29 databases, and two of them have unique quantifiers that tie together, and the rest of them don't. So, there's nothing that you can do with that data. So, artificial intelligence is just that, it's artificial intelligence, so, they know, that's machine learning, that's natural language, that's classification, there's a lot of different parts of that that are moving, but we also have to have IT, good data infrastructure, master data management, compliance, there's so many moving parts to this, that it's not just about the data anymore. - I want to ask Steve to chime in here, go ahead. - Yeah, so, we also have to change the mentality that it's not just enterprise data. There's data on the web, the biggest thing is Internet of Things, the amount of sensor data will make the current data look like chump change. So, data is moving faster, okay. And this is where the sophistication of machine learning needs to kick in, going from just mostly supervised-learning today, to unsupervised learning. And in order to really get into, as I said, big data, and credible AI does the who, what, where, when, and how, but not the why. And this is really the Holy Grail to crack, and it's actually under a new moniker, it's called explainable AI, because it moves beyond just correlation into root cause analysis. Once we have that, then you have the means to be able to tap into augmented intelligence, where humans are working with the machines. - Karen, please. - Yeah, so, one of the things, like what Carla was saying, and what a lot of us had said, I like to think of the advent of ML technologies and AI are going to help me as a data architect to love my data better, right? So, that includes protecting it, but also, when you say that 80% of the data is unsearchable, it's not just an access problem, it's that no one knows what it was, what the sovereignty was, what the metadata was, what the quality was, or why there's huge anomalies in it. So, my favorite story about this is, in the 1980s, about, I forget the exact number, but like, 8 million children disappeared out of the US in April, at April 15th. And that was when the IRS enacted a rule that, in order to have a dependent, a deduction for a dependent on your tax returns, they had to have a valid social security number, and people who had accidentally miscounted their children and over-claimed them, (laughter) over the years them, stopped doing that. Well, some days it does feel like you have eight children running around. (laughter) - Agreed. - When, when that rule came about, literally, and they're not all children, because they're dependents, but literally millions of children disappeared off the face of the earth in April, but if you were doing analytics, or AI and ML, and you don't know that this anomaly happened, I can imagine in a hundred years, someone is saying some catastrophic event happened in April, 1983. (laughter) And what caused that, was it healthcare? Was it a meteor? Was it the clown attacking them? - That's where I was going. - Right. So, those are really important things that I want to use AI and ML to help me, not only document and capture that stuff, but to provide that information to the people, the data scientists and the analysts that are using the data. - Great story, thank you. Bob, you got a thought? You got the mic, go, jump in here. - Well, yeah, I do have a thought, actually. I was talking about, what Karen was talking about. I think it's really important that, not only that we understand AI, and machine learning, and data science, but that the regular folks and companies understand that, at the basic level. Because those are the people who will ask the questions, or who know what questions to ask of the data. And if they don't have the tools, and the knowledge of how to get access to that data, or even how to pose a question, then that data is going to be less valuable, I think, to companies. And the more that everybody knows about data, even people in congress. Remember when Zuckerberg talked about? (laughter) - That was scary. - How do you make money? It's like, we all know this. But, we need to educate the masses on just basic data analytics. - We could have an hour-long panel on that. - Yeah, absolutely. - Peter, you and I were talking about, we had a couple of questions, sort of, how far can we take artificial intelligence? How far should we? You know, so that brings in to the conversation of ethics, and bias, why don't you pick it up? - Yeah, so, one of the crucial things that we all are implying is that, at some point in time, AI is going to become a feature of the operations of our homes, our businesses. And as these technologies get more powerful, and they diffuse, and know about how to use them, diffuses more broadly, and you put more options into the hands of more people, the question slowly starts to turn from can we do it, to should we do it? And, one of the issues that I introduce is that I think the difference between big data and AI, specifically, is this notion of agency. The AI will act on behalf of, perhaps you, or it will act on behalf of your business. And that conversation is not being had, today. It's being had in arguments between Elon Musk and Mark Zuckerberg, which pretty quickly get pretty boring. (laughing) At the end of the day, the real question is, should this machine, whether in concert with others, or not, be acting on behalf of me, on behalf of my business, or, and when I say on behalf of me, I'm also talking about privacy. Because Facebook is acting on behalf of me, it's not just what's going on in my home. So, the question of, can it be done? A lot of things can be done, and an increasing number of things will be able to be done. We got to start having a conversation about should it be done? - So, humans exhibit tribal behavior, they exhibit bias. Their machine's going to pick that up, go ahead, please. - Yeah, one thing that sort of tag onto agency of artificial intelligence. Every industry, every business is now about identifying information and data sources, and their appropriate sinks, and learning how to draw value out of connecting the sources with the sinks. Artificial intelligence enables you to identify those sources and sinks, and when it gets agency, it will be able to make decisions on your behalf about what data is good, what data means, and who it should be. - What actions are good. - Well, what actions are good. - And what data was used to make those actions. - Absolutely. - And was that the right data, and is there bias of data? And all the way down, all the turtles down. - So, all this, the data pedigree will be driven by the agency of artificial intelligence, and this is a big issue. - It's really fundamental to understand and educate people on, there are four fundamental types of bias, so there's, in machine learning, there's intentional bias, "Hey, we're going to make "the algorithm generate a certain outcome "regardless of what the data says." There's the source of the data itself, historical data that's trained on the models built on flawed data, the model will behave in a flawed way. There's target source, which is, for example, we know that if you pull data from a certain social network, that network itself has an inherent bias. No matter how representative you try to make the data, it's still going to have flaws in it. Or, if you pull healthcare data about, for example, African-Americans from the US healthcare system, because of societal biases, that data will always be flawed. And then there's tool bias, there's limitations to what the tools can do, and so we will intentionally exclude some kinds of data, or not use it because we don't know how to, our tools are not able to, and if we don't teach people what those biases are, they won't know to look for them, and I know. - Yeah, it's like, one of the things that we were talking about before, I mean, artificial intelligence is not going to just create itself, it's lines of code, it's input, and it spits out output. So, if it learns from these learning sets, we don't want AI to become another buzzword. We don't want everybody to be an "AR guru" that has no idea what AI is. It takes months, and months, and months for these machines to learn. These learning sets are so very important, because that input is how this machine, think of it as your child, and that's basically the way artificial intelligence is learning, like your child. You're feeding it these learning sets, and then eventually it will make its own decisions. So, we know from some of us having children that you teach them the best that you can, but then later on, when they're doing their own thing, they're really, it's like a little myna bird, they've heard everything that you've said. (laughing) Not only the things that you said to them directly, but the things that you said indirectly. - Well, there are some very good AI researchers that might disagree with that metaphor, exactly. (laughing) But, having said that, what I think is very interesting about this conversation is that this notion of bias, one of the things that fascinates me about where AI goes, are we going to find a situation where tribalism more deeply infects business? Because we know that human beings do not seek out the best information, they seek out information that reinforces their beliefs. And that happens in business today. My line of business versus your line of business, engineering versus sales, that happens today, but it happens at a planning level, and when we start talking about AI, we have to put the appropriate dampers, understand the biases, so that we don't end up with deep tribalism inside of business. Because AI could have the deleterious effect that it actually starts ripping apart organizations. - Well, input is data, and then the output is, could be a lot of things. - Could be a lot of things. - And that's where I said data equals human lives. So that we look at the case in New York where the penal system was using this artificial intelligence to make choices on people that were released from prison, and they saw that that was a miserable failure, because that people that release actually re-offended, some committed murder and other things. So, I mean, it's, it's more than what anybody really thinks. It's not just, oh, well, we'll just train the machines, and a couple of weeks later they're good, we never have to touch them again. These things have to be continuously tweaked. So, just because you built an algorithm or a model doesn't mean you're done. You got to go back later, and continue to tweak these models. - Mark, you got the mic. - Yeah, no, I think one thing we've talked a lot about the data that's collected, but what about the data that's not collected? Incomplete profiles, incomplete datasets, that's a form of bias, and sometimes that's the worst. Because they'll fill that in, right, and then you can get some bias, but there's also a real issue for that around cyber security. Logs are not always complete, things are not always done, and when things are doing that, people make assumptions based on what they've collected, not what they didn't collect. So, when they're looking at this, and they're using the AI on it, that's only on the data collected, not on that that wasn't collected. So, if something is down for a little while, and no data's collected off that, the assumption is, well, it was down, or it was impacted, or there was a breach, or whatever, it could be any of those. So, you got to, there's still this human need, there's still the need for humans to look at the data and realize that there is the bias in there, there is, we're just looking at what data was collected, and you're going to have to make your own thoughts around that, and assumptions on how to actually use that data before you go make those decisions that can impact lots of people, at a human level, enterprise's profitability, things like that. And too often, people think of AI, when it comes out of there, that's the word. Well, it's not the word. - Can I ask a question about this? - Please. - Does that mean that we shouldn't act? - It does not. - Okay. - So, where's the fine line? - Yeah, I think. - Going back to this notion of can we do it, or should we do it? Should we act? - Yeah, I think you should do it, but you should use it for what it is. It's augmenting, it's helping you, assisting you to make a valued or good decision. And hopefully it's a better decision than you would've made without it. - I think it's great, I think also, your answer's right too, that you have to iterate faster, and faster, and faster, and discover sources of information, or sources of data that you're not currently using, and, that's why this thing starts getting really important. - I think you touch on a really good point about, should you or shouldn't you? You look at Google, and you look at the data that they've been using, and some of that out there, from a digital twin perspective, is not being approved, or not authorized, and even once they've made changes, it's still floating around out there. Where do you know where it is? So, there's this dilemma of, how do you have a digital twin that you want to have, and is going to work for you, and is going to do things for you to make your life easier, to do these things, mundane tasks, whatever? But how do you also control it to do things you don't want it to do? - Ad-based business models are inherently evil. (laughing) - Well, there's incentives to appropriate our data, and so, are things like blockchain potentially going to give users the ability to control their data? We'll see. - No, I, I'm sorry, but that's actually a really important point. The idea of consensus algorithms, whether it's blockchain or not, blockchain includes games, and something along those lines, whether it's Byzantine fault tolerance, or whether it's Paxos, consensus-based algorithms are going to be really, really important. Parts of this conversation, because the data's going to be more distributed, and you're going to have more elements participating in it. And so, something that allows, especially in the machine-to-machine world, which is a lot of what we're talking about right here, you may not have blockchain, because there's no need for a sense of incentive, which is what blockchain can help provide. - And there's no middleman. - And, well, all right, but there's really, the thing that makes blockchain so powerful is it liberates new classes of applications. But for a lot of the stuff that we're talking about, you can use a very powerful consensus algorithm without having a game side, and do some really amazing things at scale. - So, looking at blockchain, that's a great thing to bring up, right. I think what's inherently wrong with the way we do things today, and the whole overall design of technology, whether it be on-prem, or off-prem, is both the lock and key is behind the same wall. Whether that wall is in a cloud, or behind a firewall. So, really, when there is an audit, or when there is a forensics, it always comes down to a sysadmin, or something else, and the system administrator will have the finger pointed at them, because it all resides, you can edit it, you can augment it, or you can do things with it that you can't really determine. Now, take, as an example, blockchain, where you've got really the source of truth. Now you can take and have the lock in one place, and the key in another place. So that's certainly going to be interesting to see how that unfolds. - So, one of the things, it's good that, we've hit a lot of buzzwords, right now, right? (laughing) AI, and ML, block. - Bingo. - We got the blockchain bingo, yeah, yeah. So, one of the things is, you also brought up, I mean, ethics and everything, and one of the things that I've noticed over the last year or so is that, as I attend briefings or demos, everyone is now claiming that their product is AI or ML-enabled, or blockchain-enabled. And when you try to get answers to the questions, what you really find out is that some things are being pushed as, because they have if-then statements somewhere in their code, and therefore that's artificial intelligence or machine learning. - [Peter] At least it's not "go-to." (laughing) - Yeah, you're that experienced as well. (laughing) So, I mean, this is part of the thing you try to do as a practitioner, as an analyst, as an influencer, is trying to, you know, the hype of it all. And recently, I attended one where they said they use blockchain, and I couldn't figure it out, and it turns out they use GUIDs to identify things, and that's not blockchain, it's an identifier. (laughing) So, one of the ethics things that I think we, as an enterprise community, have to deal with, is the over-promising of AI, and ML, and deep learning, and recognition. It's not, I don't really consider it visual recognition services if they just look for red pixels. I mean, that's not quite the same thing. Yet, this is also making things much harder for your average CIO, or worse, CFO, to understand whether they're getting any value from these technologies. - Old bottle. - Old bottle, right. - And I wonder if the data companies, like that you talked about, or the top five, I'm more concerned about their nearly, or actual $1 trillion valuations having an impact on their ability of other companies to disrupt or enter into the field more so than their data technologies. Again, we're coming to another perfect storm of the companies that have data as their asset, even though it's still not on their financial statements, which is another indicator whether it's really an asset, is that, do we need to think about the terms of AI, about whose hands it's in, and who's, like, once one large trillion-dollar company decides that you are not a profitable company, how many other companies are going to buy that data and make that decision about you? - Well, and for the first time in business history, I think, this is true, we're seeing, because of digital, because it's data, you're seeing tech companies traverse industries, get into, whether it's content, or music, or publishing, or groceries, and that's powerful, and that's awful scary. - If you're a manger, one of the things your ownership is asking you to do is to reduce asset specificities, so that their capital could be applied to more productive uses. Data reduces asset specificities. It brings into question the whole notion of vertical industry. You're absolutely right. But you know, one quick question I got for you, playing off of this is, again, it goes back to this notion of can we do it, and should we do it? I find it interesting, if you look at those top five, all data companies, but all of them are very different business models, or they can classify the two different business models. Apple is transactional, Microsoft is transactional, Google is ad-based, Facebook is ad-based, before the fake news stuff. Amazon's kind of playing it both sides. - Yeah, they're kind of all on a collision course though, aren't they? - But, well, that's what's going to be interesting. I think, at some point in time, the "can we do it, should we do it" question is, brands are going to be identified by whether or not they have gone through that process of thinking about, should we do it, and say no. Apple is clearly, for example, incorporating that into their brand. - Well, Silicon Valley, broadly defined, if I include Seattle, and maybe Armlock, not so much IBM. But they've got a dual disruption agenda, they've always disrupted horizontal tech. Now they're disrupting vertical industries. - I was actually just going to pick up on what she was talking about, we were talking about buzzword, right. So, one we haven't heard yet is voice. Voice is another big buzzword right now, when you couple that with IoT and AI, here you go, bingo, do I got three points? (laughing) Voice recognition, voice technology, so all of the smart speakers, if you think about that in the world, there are 7,000 languages being spoken, but yet if you look at Google Home, you look at Siri, you look at any of the devices, I would challenge you, it would have a lot of problem understanding my accent, and even when my British accent creeps out, or it would have trouble understanding seniors, because the way they talk, it's very different than a typical 25-year-old person living in Silicon Valley, right. So, how do we solve that, especially going forward? We're seeing voice technology is going to be so more prominent in our homes, we're going to have it in the cars, we have it in the kitchen, it does everything, it listens to everything that we are talking about, not talking about, and records it. And to your point, is it going to start making decisions on our behalf, but then my question is, how much does it actually understand us? - So, I just want one short story. Siri can't translate a word that I ask it to translate into French, because my phone's set to Canadian English, and that's not supported. So I live in a bilingual French English country, and it can't translate. - But what this is really bringing up is if you look at society, and culture, what's legal, what's ethical, changes across the years. What was right 200 years ago is not right now, and what was right 50 years ago is not right now. - It changes across countries. - It changes across countries, it changes across regions. So, what does this mean when our AI has agency? How do we make ethical AI if we don't even know how to manage the change of what's right and what's wrong in human society? - One of the most important questions we have to worry about, right? - Absolutely. - But it also says one more thing, just before we go on. It also says that the issue of economies of scale, in the cloud. - Yes. - Are going to be strongly impacted, not just by how big you can build your data centers, but some of those regulatory issues that are going to influence strongly what constitutes good experience, good law, good acting on my behalf, agency. - And one thing that's underappreciated in the marketplace right now is the impact of data sovereignty, if you get back to data, countries are now recognizing the importance of managing that data, and they're implementing data sovereignty rules. Everyone talks about California issuing a new law that's aligned with GDPR, and you know what that meant. There are 30 other states in the United States alone that are modifying their laws to address this issue. - Steve. - So, um, so, we got a number of years, no matter what Ray Kurzweil says, until we get to artificial general intelligence. - The singularity's not so near? (laughing) - You know that he's changed the date over the last 10 years. - I did know it. - Quite a bit. And I don't even prognosticate where it's going to be. But really, where we're at right now, I keep coming back to, is that's why augmented intelligence is really going to be the new rage, humans working with machines. One of the hot topics, and the reason I chose to speak about it is, is the future of work. I don't care if you're a millennial, mid-career, or a baby boomer, people are paranoid. As machines get smarter, if your job is routine cognitive, yes, you have a higher propensity to be automated. So, this really shifts a number of things. A, you have to be a lifelong learner, you've got to learn new skillsets. And the dynamics are changing fast. Now, this is also a great equalizer for emerging startups, and even in SMBs. As the AI improves, they can become more nimble. So back to your point regarding colossal trillion dollar, wait a second, there's going to be quite a sea change going on right now, and regarding demographics, in 2020, millennials take over as the majority of the workforce, by 2025 it's 75%. - Great news. (laughing) - As a baby boomer, I try my damnedest to stay relevant. - Yeah, surround yourself with millennials is the takeaway there. - Or retire. (laughs) - Not yet. - One thing I think, this goes back to what Karen was saying, if you want a basic standard to put around the stuff, look at the old ISO 38500 framework. Business strategy, technology strategy. You have risk, compliance, change management, operations, and most importantly, the balance sheet in the financials. AI and what Tony was saying, digital transformation, if it's of meaning, it belongs on a balance sheet, and should factor into how you value your company. All the cyber security, and all of the compliance, and all of the regulation, is all stuff, this framework exists, so look it up, and every time you start some kind of new machine learning project, or data sense project, say, have we checked the box on each of these standards that's within this machine? And if you haven't, maybe slow down and do your homework. - To see a day when data is going to be valued on the balance sheet. - It is. - It's already valued as part of the current, but it's good will. - Certainly market value, as we were just talking about. - Well, we're talking about all of the companies that have opted in, right. There's tens of thousands of small businesses just in this region alone that are opt-out. They're small family businesses, or businesses that really aren't even technology-aware. But data's being collected about them, it's being on Yelp, they're being rated, they're being reviewed, the success to their business is out of their hands. And I think what's really going to be interesting is, you look at the big data, you look at AI, you look at things like that, blockchain may even be a potential for some of that, because of mutability, but it's when all of those businesses, when the technology becomes a cost, it's cost-prohibitive now, for a lot of them, or they just don't want to do it, and they're proudly opt-out. In fact, we talked about that last night at dinner. But when they opt-in, the company that can do that, and can reach out to them in a way that is economically feasible, and bring them back in, where they control their data, where they control their information, and they do it in such a way where it helps them build their business, and it may be a generational business that's been passed on. Those kind of things are going to make a big impact, not only on the cloud, but the data being stored in the cloud, the AI, the applications that you talked about earlier, we talked about that. And that's where this bias, and some of these other things are going to have a tremendous impact if they're not dealt with now, at least ethically. - Well, I feel like we just got started, we're out of time. Time for a couple more comments, and then officially we have to wrap up. - Yeah, I had one thing to say, I mean, really, Henry Ford, and the creation of the automobile, back in the early 1900s, changed everything, because now we're no longer stuck in the country, we can get away from our parents, we can date without grandma and grandpa setting on the porch with us. (laughing) We can take long trips, so now we're looked at, we've sprawled out, we're not all living in the country anymore, and it changed America. So, AI has that same capabilities, it will automate mundane routine tasks that nobody wanted to do anyway. So, a lot of that will change things, but it's not going to be any different than the way things changed in the early 1900s. - It's like you were saying, constant reinvention. - I think that's a great point, let me make one observation on that. Every period of significant industrial change was preceded by the formation, a period of formation of new assets that nobody knew what to do with. Whether it was, what do we do, you know, industrial manufacturing, it was row houses with long shafts tied to an engine that was coal-fired, and drove a bunch of looms. Same thing, railroads, large factories for Henry Ford, before he figured out how to do an information-based notion of mass production. This is the period of asset formation for the next generation of social structures. - Those ship-makers are going to be all over these cars, I mean, you're going to have augmented reality right there, on your windshield. - Karen, bring it home. Give us the drop-the-mic moment. (laughing) - No pressure. - Your AV guys are not happy with that. So, I think the, it all comes down to, it's a people problem, a challenge, let's say that. The whole AI ML thing, people, it's a legal compliance thing. Enterprises are going to struggle with trying to meet five billion different types of compliance rules around data and its uses, about enforcement, because ROI is going to make risk of incarceration as well as return on investment, and we'll have to manage both of those. I think businesses are struggling with a lot of this complexity, and you just opened a whole bunch of questions that we didn't really have solid, "Oh, you can fix it by doing this." So, it's important that we think of this new world of data focus, data-driven, everything like that, is that the entire IT and business community needs to realize that focusing on data means we have to change how we do things and how we think about it, but we also have some of the same old challenges there. - Well, I have a feeling we're going to be talking about this for quite some time. What a great way to wrap up CUBE NYC here, our third day of activities down here at 37 Pillars, or Mercantile 37. Thank you all so much for joining us today. - Thank you. - Really, wonderful insights, really appreciate it, now, all this content is going to be available on theCUBE.net. We are exposing our video cloud, and our video search engine, so you'll be able to search our entire corpus of data. I can't wait to start searching and clipping up this session. Again, thank you so much, and thank you for watching. We'll see you next time.

>> Narrator: From Hell's Kitchen in New York City, it's theCUBE, on the ground at Serverlessconf brought to you by SiliconAngle Media >> Hi, I'm Stu Miniman, here with theCUBE at Serverless Conference in New York City, Hell's Kitchen. Happy to have with me, first time guest on the program Sam Kroonenburg, we had your brother on the program at the AWS Summit not far from here, at the Javits Center in New York City, but you're also one of the co-founders its the two brothers for A Cloud Guru. Thanks so much for joining me, and thank you for allowing us to come get some phenomenal content here. >> Yeah, no problem. Thank you for coming for the conference today. >> Alright, so Sam, take me back, you know, we talked to your brother a little bit about, well it was an interesting story, he said actually I got turned down for a job from Amazon and ended up creating a training company. But you built this and you built it on Serverless. >> I did yeah. >> So walk us through a little bit the thought process, the timing, you know, aren't you a little bit ahead of your time on that? >> Yeah, it was mid 2015, it was a strange time. We decided we wanted to build this school, this online learning platform, but the challenge we had was that we didn't have a lot of time, we both had families, kids, you know, mortgages, financial commitments. Basically I had four weeks. I had four weeks of leave owing to me, from my employer at the time. My wife and I had been planning this big family holiday with the kids for years and we were about to take it, and I remember having this phone call with Ryan and we were talking about how there were these people taking these online courses and they were really liking them. And we thought, what if we could build this school to teach people cloud computing. It was such a buzz and we just thought, there's something in this. But the challenge was the timing. I remember my wife turned to me and she said, "Look you've got to do it, we'll cancel the holiday, "take the four weeks and give it a try." So that's what we did, we actually flew down to live with Aaron, my in-laws and help look after the kids and I locked myself in a bedroom for four weeks and tried to build an online school. And that was there was no epiphany to go Serverless there was no grand plan. It was, we had a constraint, which was time. I had no time to build this thing. And so ended up using some of the latest technologies like AWS Lambda, API Gateway, a whole bunch of Serverless technologies because I saw that they would help me build this faster. And I could get something to market in the four weeks that I had. I actually spent the first couple of days trying to skin and configure Moodle, the learning management system and I tore my hair out and yeah, ended up putting this thing together with Serverless technologies. >> Ryan just walked by-- >> Oh, there he is. >> It's a llama unicorn with a cat or something like that. >> I'm going to put in the background. >> In the back of our video. Sam, what's your brother doing here? >> He's always trying to troll me. >> So talk to us, you know one of the things the maturation, kind of the speed of change in the industry for new technologies is just so fast these days. Take us through from those early days to you know Serverless today. What's your experience been? What would you say to people that look at this technology? >> I think it's a lot easier to get into now than it was two years ago. The ecosystem has grown around it, the core technologies are pretty much the same as they were two years ago, function as a service, execute functions in the cloud very similar, but the tooling around it, the ecosystem around it has grown. There's great deployment tools, orchestration systems that have come along. It's a lot easier to just get in now and early on, when we started we had to roll a lot of things ourselves, which took a lot of time, and that's what you're trying to stop, is losing time. Yeah, so there's that and the community has really grown, there's a lot of support in the community now. >> So if you had to do it all over, you could have done it in a weekend, rather than the four weeks. >> Yeah, instead of the four weeks. >> Yeah, I mean what's-- >> That's the interesting thing about what happened to us, we would not exist, our business would not exist if it wasn't for Serverless technologies. I literally couldn't, we could not have, built that school. It's not like it was the most amazing school when we launched it, but it was enough. It was just enough to get people using it, to get to market, to start to build a business around it. >> Alright, talk to me about this event. So, its the 5th Serverlessconf, not unheard of a company that does training to get involved with physical events, 'cause you bring them together, you know, what's the thought process, talk to us a little bit about that journey and this event itself. >> Yeah, I mean, a lot of this is organic for us. We built, it was early last year, you know we're part of the Serverless communities, a lot of pioneering going on here, a lot of people facing the same challenges. And we thought, well there's no event to bring all of these people together. And there's a lot of very fast pace of change here, a lot of rapid ideation and new technologies. Let's bring everyone together and see what we can do. That's what we did with Serverlessconf. We've never run a conference before, we just hired a warehouse in Brooklyn, a bunch of Australians and British guys coming over and we just invited a bunch of people on Twitter and 250 people turned out to the first one. It just got bigger and bigger from there. So this is actually the 5th Serverlessconf now. >> Well, its a hot week again, so we appreciate that the air conditioning works at this one. >> Yes, we have air conditioning at this one. >> 460 people here, you brought in some great speakers, we had a number of them on our program this week, speak to us, I mean you've got sponsors here, you've got good speakers, give us some of the highlights. >> We've got all of the main Cloud vendors are here, Google, IBM, Microsoft, Amazon and it's actually the product teams who build this stuff. That's what I love about this event, it's actually the people who build it. It's vendor neutral, it's really cool. You get great thought leaders from the community, Simon Wardley was a highlight this morning, his talk on Value Chain Mapping and Strategy was really interesting. Randall Hunt from AWS X Space X, talking about the continuous integration process when building rockets. Space X was absolutely fascinating and what bugs in production mean when you're building a rocket. It means the rocket blows up. Really interesting variety of talks from those tooling providers, companies like us who are just building on Serverless and then Serverless tooling companies and vendors. Really fascinating. >> Alright, Sam what should we be looking for in the future from Serverless and from A Cloud Guru? >> We're going to be doing a whole lot more Serverless content. You're going to see a lot of really interesting new content through our site, a lot of teaching on Serverless, we're going to be doing more Serverless Conferences. You'll see a lot from us, not just us, but from the wider community who come to the conference, who we know well, a lot of the experts, we're going to be doing a lot of work with those people. >> Well Sam Kroonenburg, really appreciate you joining us, appreciate the media sponsorship to allow theCube to come get some great content and share it with our communities, hope to see you at many more events in the future. >> Thank you for coming. >> Thank you so much. Sam Kroonenburg, I'm Stu Miniman. Thank you for watching theCUBE. (upbeat music)

>> Announcer: From Hell's Kitchen in New York City. It's theCUBE, on the ground, at Serverlessconf. Brought to you by SiliconANGLE Media. >> Hi, I'm Stu Miniman, here with theCUBE at Serverlessconf 2017 in New York City, Hell's Kitchen, actually, happy to welcome to the program, hard to believe, someone, as far as I can tell, we've never had on the program-- >> Yeah, I don't think so. >> But I've known for a long time, actually been drinking with him in Hell's Kitchen before, so Aneel Lakhani, thanks so much for letting me interview you. Your current position is vice president of marketing at Honeycomb.io >> Correct. >> Do you just call it Honeycomb or-- >> We just call it Honeycomb. >> Alright. So, Aneel, how are you doing? Tell us a little bit about your background, but keep it short and what gets you involved in the whole serverless ecosystem? >> Yeah, sure. So, about me, I've been in tech for a little over 20 years now, started out as an engineer, moved through a bunch of systems roles, architecture roles, and product roles, and now I run marketing at start-ups, which is what I've been doing for the last half decade or so. >> I think back to when Amazon announced Lambda, everybody's like, "Ooh, it's cool, what is it? "How do I use it?" Things like that. One of the things I've heard out of this event, this week, is tooling, monitoring, understanding, digging into it, which really falls into Honeycomb space. >> Yeah, I mean, it sort of does. I mean, at Honeycomb we do what we call observability, which is something a bit larger than just monitoring, right. It's really getting to the point where you can develop an understanding of what your services and what your code do in real life under real load with real users. >> Speaking of John Willis, about what is the role of operations when I don't own the infrastructure, I have to trust someone else to do it. So, bring us in there a little bit, what are some of the challenges people are having, how do they help when they're leveraging? >> Yeah, so something that's very clear about serverless approaches to building things, and especially if you're using something like Lambda, is that, as a software engineer, who writes a function, you are 100% responsible for all of your operations at that point, because the ops people for your stack are behind an API. You are on the other side of that API, and what they do is effectively a black box, which means you have to not only understand what your thing does, you have to understand what they do and how they do it, and it's some means of accessing both those bits of data. So you get what Amazon tells you for Lambda, or what any of the other providers tell you for their functions, but you also have to then understand how your code performs on that specific provider, which means you have to do things like wrap your functions in timers and emit events which go into Kinesis, or wherever else, so that you can track what's going on. >> Yeah, one of the problems, of course, any time you have any layer of abstraction is, if things go wrong, how do you get the expertise to know, how do you get in there, is this even worse in serverless? >> Yes and no, I mean, it depends on how much faith you have in your provider, right? So, one of the companies here put up a chart that shows you the performance, on average, of the call response time for the functions for all of the providers that provide serverless infrastructure. And they're not even remotely consistent. They're not consistent within even a few percentiles. In other words, if you care about performance, and you care about predictability for your function, it's basically impossible to get that from any given provider. >> Alright, so, talk to us, what are you hearing from users these days, what's exciting you in this space? >> Yeah, so what we hear from our users, anyway, at Honeycomb, who are using Lambda, and using serverless functions, is that the ability for them to get visibility into how a function performs is basically the highest priority outside of writing a function itself. Because they don't know what's happening below them, they don't know all the resource allocations at any given point in time by the provider, so the thing they have to go on, for the rest of the black box, is how their own function performs, which means they need the ability to take any given function and either decompose it into parts, which have their own events or metrics or telemetry that they emit, or they need to do that to the entire function from end-to-end. So basically have a concept of, this is an old concept for us, which is an end-to-end check, right? I want to know what happens when a point that I touch with a sim until my entire set of functions are complete at the end. >> Yeah, we're going back to like an IP ping, right? >> That's right, yeah, effectively. >> Today, Honeycomb, do you only support Lambda, do you support some of the other serverless frameworks that are out there? >> So, we are agnostic. So, basically, the way Honeycomb works is that our users instrument their code, and we're not service-only, it could be any code running anywhere, and they emit data, and that data is in the form of structured events, those structured events are consumed by Honeycomb, and then Honeycomb turns around and lets you do fast analysis against it. >> Yeah, you've got a lot of background of, "How do we leverage the knowledge of the crowd?" >> Yeah. >> So many times it's what are people finding when they're really getting involved here, you're tooling and others, what mistakes are they making, how can they get better, faster at what they're doing? >> Yeah, a common mistake that people make is not thinking about what is and is not blocking within their functions, and not understanding the threading model of the underlying stack, and when they should spin up additional functions and split up work, versus when they shouldn't, and the only way to understand that is, one, to read all the damn docs, and two to experiment. >> Yeah. What about the maturity of serverless? There've been a lot of discussions here. I had Mark from Trendade on, we talked about security, and the like, but what do you see, kind of in the maturation cycle, of serverless, anything you've heard, or still things that are looking to get fixed even more? >> Maturity isn't the word that I want to use here, I think it's more interesting to think of it in terms of breadth of capabilities, right? So, all of the serverless offerings for all of the vendors have limitations on either the programming languages you can use or the nature of the functions that can be run or the research allocation you can have. I think there's not a lot of maturity that we're going to see from the vendors other than more consistent performance, what we are going to see maturity in is, from the users' standpoint, of how they construct things. >> Yeah. Any data you can share is just how prevalent serverless is out there in the wild, you know, what's the typical use taste, typical customer kind of order of magnitude, how many people are doing it, and therefore driving discussions? >> Yeah, I have no idea. >> You have no idea about this. >> What I do know is, in our user base, we have some significant users of Honeycomb who are 100% run on Amazon Lambda, but that's my tiny, little sample size. >> Okay, want to give you the final word, serverless conference and serverless in general, what's your take today, what should people be looking at in the next six or 12 months? >> Yeah, so I more-or-less agree with Simon Wardley about this, which is, effectively this is a way for Amazon to eat most of the tech ecosystem, assuming people become dependent on it. >> Alright, well, I always say with theCUBE we like to take those hallway conversations, someone that I've had many hallway conversations with, and over the Twitters, and other ways, it's great to catch up with you, Aneel Lakhani, thanks so much for joining us >> Thank you so much. >> I'm Stu Miniman and thanks for watching theCUBE.

>> Narrator: From Hell's Kitchen in New York City, it's The Cube on the ground at Serverlessconf. Brought to you by SilliconANGLE Media. >> Hi, I'm Stu Miniman with The Cube, and we're here are Serverlessconf in Hell's Kitchen New York City, really happy to welcome to the program, another one of the keynote speakers. Ben Kehoe, who's the Cloud Robotics research scientist at iRobot. >> Yeah. >> Ben, great to see you. >> Great to see you too. >> All right, so tell us a little bit about how you got involved with Serverless. >> Yeah, I mean it all started, I was a grad student in robotics, and I started thinking about, you know, we have all these robotics algorithms. And as the cloud can enable robots to do more and better things, how do we help turn those robotics algorithms into web services. And I didn't get very far in that, right towards the end of my PHD, and then that was 2014, LAMBDA was released, and it was like hey, that looks like it does the kind of thing that I was thinking about that we needed. So then I joined iRobot, and we were developing a cloud solution, a cloud application for our connected robots and apps, and to help us scale that to stay lean. Serverless was the right choice, and we've been doing that since 2015. >> Yeah, so Ben, what is it about Serverless that made it a fit for this? You know, I think about, doesn't their responsiveness, performance, latency if I have to go >> Yeah. >> up to the cloud and back like that way. I think some of this needs to kind of live locally. And some that goes there, maybe you can just briefly tease through some of those dynamics for us. >> Yeah, when you're talking about robots, you definitely have to keep things local. You want a robot to be responsive to its environment. You want, that even if its cloud connection disappears, that it can still accomplish all of its tasks. So it's always a mix of keeping it as a timeless robot that is enabled to do better things through the cloud, in terms of additional computational power, or accessing libraries of information to help it understand its world better. And of course, when one robot learns something, all robots can benefit from that experience. >> Excellent, so this is the first step for Skynet is what you're saying, right? >> Could be. >> All right, bring us in a little bit. Your keynote, what were you looking to share? You know, some of the key points. >> Yeah, I think in the talks that I've given at Serverlessconf, they tend to be as much as I am enthusiastic about Serverless, fully bodying, I try and pull us back a little bit to say, "What are we still missing? "What's not here yet? "Where do we need to go?" And so I had some frowny face emoji in my talk about event driven programming, event driven Serverless, and Serverless without event driven programming. Now we're still, you know, we have areas to improve in each one of those. And then that transitioned really into, "How do we start bringing in people who "are just starting into Serverless?" Larger organizations, more traditional architectures, and people who are experienced with that, and understand traditional architectures well. How do we get them on board with Serverless? And so that starts with just the gateway drug, which is infrastructure automation at the edges of their application, taking scripts that they run from developer machines with Cron jobs, and moving those into a function that's triggered by some cloud event. And then from there, starting to bring them over in terms of you can reduce your costs by eliminating idle resources. You can start to simplify and strengthen by refactoring some of that. And then once you really get them thinking about, "Oh, this is really working for the things "that we're doing." New features will start to be developed. Serverless native or event driven native. And then sort of at the end of the talk, the key is that because Serverless architectures look different from traditional architectures, there's something called Conway's law that says, "The design of your application will follow "the communication patterns in your organization." >> Stu: Right. >> And so you have to sort of flip that around to say, "Well if our design is changing, then we have "to make our organization change as well." >> Right, does that mean we're going to have, micro-employees you know? Instead of micro services we have, you know, employees that we hire them, and then we fire them pretty quick when we don't need them, or? >> I hope not. >> Yeah. >> I hope not. >> (crosstalk) that that's the part time, the uber's >> Yes. >> nation of the workforce. >> Yes. That would be, I think an inefficient way of going about it. >> Yeah. >> But I think we do need to reset expectations around what we have control over, and what we don't, because when you're on a traditional architecture with servers, you can reach in and fix problems that you have. And recognizing that when you're running on functions as a service platform, and using managed services, that when the provider has some sort of incident, you're out of control of that. It's a very uncomfortable place to be of not being in control of your own destiny, even though when you look at the big picture, that's going to happen less often, then if you were doing it yourself. >> Stu: Yeah. >> And so that's making sure that the mindset inside the organization, and the way that people communicate, is appropriately tuned to that sort of new paradigm. >> Okay, yeah. Ben, some of those frowny faces, what are things that the community is working on that you're hopeful for? What are some of the areas that we need for the maturation of this space? >> Yeah, I think something that I talked about previously that's coming around, is monitoring. So there's much more tools out there to monitor the infrastructure to know what's going on inside these functions and these managed services. And there's now some security analysis tools that are coming out, that some of these people are present here. And that was a big aspect that I've harped on for a long time of... We have a lot of mature traditional tools, that will do network analysis of your servers. Well it's like, "I don't have any servers." And those vendors then say, "Well, we can't help you." And there's static code analysis vendors who say we look at your whole application, and the flows inside it. And we say, well most of my application exists outside of code that I've written. I just write little bits, that glue it together in the way that my business works. And they say, "Oh, well we can't help you." >> Yeah. It reminds me, I think you know for so many years, people were really excited about how they could build their infrastructure. >> Yeah. >> And now they look to environments, well I can get out of that. So it caught my eye. You know, you put out on twitter, said "Maybe we need to have, you know, my next talk will be, "Work dumber not harder." Maybe explain that a little bit. >> Yeah, so I think, >> Yeah. >> I've been thinking about, you know, with some of the talks here about how it's not building it yourself. That in some ways, there's not invented here syndrome. And we kind of want to go a little bit down the road of invented here syndrome, of if you're building something that is not business logic, you're probably ideally thinking, "Maybe I shouldn't be doing this." So turning it into, I don't want to have to be clever in setting up my architecture, because being clever and like writing, it's always interesting to do, right? When you're developing, you're solving a computer science problem. But often that mean you're not delivering business value. And so, in Paul Johnson's talk, he was talking about the kind of people he looks like. What the kind of people he looks for, look like. >> Yeah. >> And he was saying, you know, "It's people "who want to get stuff out the door. "And who think about good enough." And I think that's really the thing of, how do we, when the people you hire are people who just want to ship features, they're going to say, "I can pull together services to do that "without having to actually solve any hard problems." And that means that you're delivering value, and you're operating more in your business space then in a technology space." >> All right, Ben I want to give you the final word. >> Thank you. >> You know, only 460 people here, which is good growth for the show, but a lot of people out there that are still learning about Serverless, what tips do you give them? You know, first steps to get involved, get involved with the community, (mumbles) some early wins they can have? >> I think there's a couple of things. There is training out there, there's blogs. There's twitter. Ask questions. You know, ping me on twitter if you wonder about something. And there's a Serverless slack that's very active, and if you ask basically anybody, the link is floating around. >> All right, well Ben Kehoe, thanks so much. Great to meet you, and thanks for sharing in this community. >> Yeah, thanks for having me. >> And our community, I'm Stu Miniman and thanks for watching The Cube. (upbeat, exciting music bumper)

>> Announcer: From Hell's Kitchen in New York City, it's theCUBE, on the ground at Serverlessconf. Brought to you by Silicon Angle Media. >> Hi, I'm Stu Miniman with theCUBE, here at Serverless Conference in Hell's Kitchen in New York City. Happy to welcome back to the program. keynote speaker at the event, and a guest that we've had on a couple times before, John Willis, who's the vice president of DevOps and digital practices at Eastray Technologies. John. >> In Hell's Kitchen. >> Stu: In Hell's Kitchen, and go Yankees. >> Yeah, man. I was at the game last night, the other night. Yeah. You'll see tonight. Yeah. Thank you. Glad to be here. >> Great to see you. So look, you've been talking to audiences about DevOps for as long as I can remember, as long as I've known you, definitely. Tell us, what's so important about serverless and how that fits into the world of the developer these days. >> Yeah, I mean, my interest, you know, I was invited to do a keynote, and my interest is to break down the tribal nature of new things. And I sound like a hypocrite because I'm the DevOps tribe, but I prefer to stop calling it DevOps, because there are super patterns that exist, and as I watch serverless, I spend a lot of time having these conversations around that yeah, we don't need that DevOps anymore, because we got serverless. It was the same reason like we didn't need any of the infrastructure stuff because we got cloud. And like, we keep throwing the baby out with the bathwater, and my presentation this morning was like, it's not about the technology, stupid. Like the principles of business value, how you understand value stream, how you inject the governance, the policy, the security, the values and the outcomes that you want. I know those sound like platitudes, like I get a sense that we're making the same mistake over again, and hey, sorry folks, Serverless is just another form of compute. Sorry to get you all wound up and then let you down. It's just compute, folks. And so all the core principles that we've really learned about high-performance organizations apply, they apply differently. Monitoring is differently. How do we deliver? But the principles stay the same. And that was my core message today. >> Yeah, no, very passionate, definitely came through in the keynote. I just have to ask you just on the tech for a second, I mean you were heavily involved in containers, you were part of a company that got acquired by Docker, you were a big proponent of unikernels, now it's serverless, how do you kind of paint that picture >> I think it's amazing tech, and more these days. So I left Docker and I'm going back to something I did 10 years ago, which is kind of consulting but transformation type consulting. It sounds platitudish, but like, I'm back in the mode of looking at things at bigger scale. How do you change an organization to think differently about things? So I've kind of taken a little bit of my tech hat off. I mean, I love containers and minimal delivery, right, I've been yacking about that for like the last two or three years, right? About how minimal delivery models work. And serverless is like, amazing too, like unikernels was an interesting model of function as a service. I think serverless will eat up a good portion, you know I've said this, and I don't know, I may have to modify it. You know, I would say four years ago, three years ago, and you guys been a big part of this discussion. The world went to most companies would say we're a cloud-first organization. I've been saying for the last couple of years, I think most organizations should now thinking that they're a container-first organization. So that doesn't say everything, it just means, and I think the world now should be kind of still container first, and I know that might sound horrible to serverless people, but then look at serverless functions as a place where it fits in the architecture, repeatability, and containers. And there's actually kind of a.. >> Is that just from a maturity standpoint, you know, containers a little bit more mature than serverless? >> I don't know that it's, I think there are like, there are models of architecture, right, and I don't know that, I mean I know there's a lot of successful startups in certain value streams and enterprises that are all serverless. I know a couple of friends that have built complete infrastructure on Amazon Lambda. It works. I just don't know that all value stream delivery of services will go complete serverless. I'm pretty certain that today, almost all applications can run on containers. So I'm not creating a division of war. I'm just saying that I think, and I could be dead wrong on this, but I think in this future like placeholder where we're container first, it's going to be, give me an exception of why it can't be containers left, like it has to be cloud, or it has to be bare metal, or it has to be (mumbles) and the right side is about mapping reusable functionality in functions. So I think you have like a container-first world assumes that smart architecture mandates repeatable functions in a function-like world. Does that make sense? >> Yeah, it does. So I think back on my career, there's so many times we said like, oh, we've got this new way to really simplify the environment and get rid of things you don't need to worry about. You know, I lived through the whole virtualization, oh wait, networking storage took us a decade to fix that. >> Yeah, yeah, yeah, yeah. >> Containers, oh we're going to just focus on the application. Oh wait, networking really important, you worked on a whole company focused specifically on that. >> DevOps for networking, yeah. >> Serverless, the question is, what's the rule of operations when it comes to serverless? >> Again, that's my thoughts on serverless and if it ain't right that's secondary to my real passion right now, which is when I hear the word NoOps for serverless, I cringe. Like this idea that you don't... I mean it's different. Do you need observability and telemetry in a serverless world? I ask you. Of course you do. Do you need to have repeatable patterns of delivery to make sure you don't have vulnerabilities in your code? Of course you do. That's Ops folks. And it's about supply chain and building repeatable, structured delivery with all the gates and the checks and the units, and none of that I believe goes away with serverless. Just like it didn't go away with cloud, just the way it didn't go with virtualization, right? So I think you know, we make a big mistake to think serverless means we don't need operations now. Does it mean that our providers, we have a different relationship with our providers? We don't own the server anymore. So we can't run detrace or those kind of things in that environment. But we still own the service. So who's the site reliability engineer for the service that's running on Lambda? Or functions of serverless, right? If it ain't, I mean if you don't got one, like you're going to have a bad service. >> Yeah, what are you hearing organizationally, what's happening in companies that you're talking to? You know, I was a at a show recently, I think it was Kelsey Hightower I think, it was like DevOps is a given at this point. So do you see that, you know, where's the line from what you've seen? >> Well the curse and the blessing of DevOps, the curse is we've never had a clear definition of it. I say we, you know, everybody, but. And the blessing is we've never had a clear definition. Like it's always emerged. And the problem is, I will tell you what my definition of DevOps is, it has really very little to do with technology. It has to do with human capital and how you create high-performing organizations and the principles and practices that lead to that. The DevOps handbook, if you will, is a lot about, that I co-authored with Gene and Patrick and Jez. Those things, that's my definition of DevOps, but the problem is, when you hear people have discussion about DevOps in lieu of a good definition, you can't really get upset when somebody thinks DevOps is like Jenkins and Sheffer Puppet and Ansable, and like oh no, you're wrong, right, like that's their view. So the problem that you run into then is, if your definition is that it's pure technology and it's tied to kind of cloud, and it's something like infrastructure is code, then in your world and your definition, serverless is going to make all that obsolete, or a good portion obsolete. But if your definition is more about how you create patterns and practices around humans who deliver services a certain way, then nothing about serverless makes any of that obsolete. >> All right, Jon, want to give you final word. What do you think people, that you know, just hearing about serverless first time, where do they start, what kind of things should they look at, or you know, if there's other things you think they should probably look at first? >> You know, I think you're asking the wrong guy for that really. I think there's far better people that you've interviewed take care of that. I mean I would go with Peters Brook, the founder of this conference. That was a book I read, he gave me a copy, it made sense to me, I was able to do some labs and then you know, as they say, the rest, Bob's your uncle, you know, there's a ton of stuff out there to figure out how to navigate. >> Anything, any commentary you'd make on the community for here, a couple of people just you know, it's new but very vibrant, reminds me a lot of the emerging tech where, you know, a lot of help from the community, it's pretty easy to get started. >> So yeah, so in the technology, yes. A lot of vendors, a lot of good stuff, great conversations, and I was actually pleasantly surprised there was less discussion about NoOps or you don't need operations, and I got kind of a little bit of a cheer when I mentioned that this morning. So it seems like there are some good lessons learned that I think the message loud and clear is that operations still exist, it just has to be thought about. The keynote yesterday, the gentleman in the keynote yesterday said, day one, closing keynote, said serverless things are different, in some case easier, but harder in other things, and that was through a cloud. Cloud was much easier from getting infrastructure but we ran into a whole lot of operational issues around how to match this cloud to scale. So serverless is easy to create a function, get it set up, cost-effective, but we're starting to learn all of the complex operational issues of MTTR, how do you restore stuff, what does SRE look like, I mean this is why we get paid the big bucks, dammit man. >> All right, John Willis, always a pleasure to catch up with you. I'm Stu Miniman, thank you so much for watching theCUBE.

>> Narrator: From Hell's Kitchen in New York City, it's theCUBE. On the ground at Serverlessconf. Brought to you by SiliconANGLE Media. >> Hi I'm Stu Miniman, here with theCUBE at Serverlessconf in New York City, really excited to have on the program one of the keynote speakers and a first time guest on theCUBE, it's someone I've know through the interwebs and have read his stuff for many years, Simon Wardley who's a researcher with a leading edge firm, Simon, great to see you. Thanks so much for joining us. >> Thank you ever so much for inviting me. It's a delight to be here. >> Alright, so my understanding is thanks to this event, you've reached a lifelong career goal. You're now a Sith Lord? (laughing) >> Well, somebody basically took a quote of mine and put it on a Star Wars poster with The Empire at the bottom, so yes, it is absolutely there you are, I am a Sith Lord, so delightful. >> The quote was that Serverless will just fundamentally change the architecture of how we build things. Something along those lines, I believe. >> Absolutely, yes. >> Alright, so let's start there. There are so many, come on, we all got really excited when containers came out. We're going to talk to John Willis >> You did. (laughing) >> We're going to talk about unikernels. The industry as a whole, there's frothiness and buzz >> Okay. >> So Serverless, you know, how's it different? How's it the same? Why's it so important from your standpoint? >> So, really good questions. So, to explain that question, we have to start off with a subject that is dear to my heart which is mapping. So when we look at the value chain of any organization, the components in that value chain are evolving and they evolve from the genesis, the novel and new to custom built examples and eventually products and rental services and then commodity and utility services. And that process is driven by supply and demand competition. It happens not only to activities, but to practice and data, but we give them different terms. They have all of the same characteristics as when they evolve. Now, when you look at that evolving environment, what you discover is there are two basic forms of disruption. There is the highly unpredictable form, which either occurs due to the appearance of something novel and new, which we don't know what it's going to impact or to product substitution. So that's the Nokia versus Apple, sort of battle, you don't know which way it's going to go until after the battle. And there is a second form of disruption, which is much more anticipatable or predictable and that is the product to utility change. So we know that when things evolve from product to utility we're going to see a rapid period of change and then there's a punctuated equilibrium. Explotion of higher order systems. We're going to see co-evolution of practice, disruption of past companies stuck behind inertia barriers. Yes it's going to be a bad efficiency, no we're not going to save any money 'cause we're just going to do more stuff with it and we're going to have all these new things as well. And we can anticipate that in advance. So when you start looking at value chains of organization, it's always the shift from product to commodity and utility which makes the big transformation in industry. And so one of them was compute. Shifting from products, as in servers, to utility as in cloud. Unfortunately dreadful term, cloud, an awful word, you know it's not a wispy thing up in the sky, it is something very specific, the shift from compute to utility. >> Would you put virtualization along that continuum? >> Okay, so virtualization was one of the underlying components, which actually helped with that happen. >> Yes. >> And so you've also got the explosion of practices around that co-evolution of practice, things like DevOps. Well, the same transition is now happening in the platform space. So, we're moving away from a product stack, things like, LAMP and .NET, to much more utility-based code execution environments. And that's what we're getting with Lambda. And we're going to see an explosion of new things built on top, inertia barriers, companies stuck behind, they'll die off, It'll be a rapid change punctuated equilibrium. You'll get all sorts of new things built. So we're going through that big transformation. Now, these transformations have been going on for about 300 years, some of them impact micro scale economics, some macro, the biggest we call ages. And that all depends upon how widespread that component is in other value chains, so when we're talking about software, we're talking about a component which is in almost all other value chains, it's shifting from product to utility, massive change, highly predictable. This is what Serverless is about. So, will it change everything? Absolutely it will. >> Alright, so Simon, I'm wondering, if you've mapped out for Serverless, where's the land of economic expection, the land of happiness and the land of despair? (laughing) >> Well, okay, happiness, despair and expectation? >> Yes. >> Okay interesting one. So the land of despair will be getting stuck behind the inertia barriers, dismissing it, saying it's not going to impact, it's not going to impact, no, no, because there's a punctuated equilibrium, it'll surprise you because it's an exponential growth, so you'll think you've got loads and loads of time and 10 years from now, you're like, be panicking, oh my gosh, it's impacting, I can't get the skills for people to help me do the transformation. My entire industry and business model is starting to disappear, so that is the land of despair that's coming to people, that's easy to defend against because most people can't see the environment. They're going to just walk straight into that one. The land of happiness. Well, obviously other than being the utility providers who'll be extremely happy about the growth of their industry, another area of happiness will be some of the novel and new things built on top. So, we're bound to see the, sort, of, one person, two person company who builds a fuction which is sold through something like the marketplace and everybody uses and they sell it for a billion. So, we'll get the two person billion dollar company and I'm sure that will make them delightfully happy. So, that's despair, happiness, also inflated expectations. So one of the big lies will be, Serverless is going to save me money in terms of reducing my IT budget. I'm afraid not. This is Jevons Paradox, this is being going on since 1865. All that's going to happen is yes, it becomes more efficient but we'll do more stuff because we're in competition so we'll spend exactly the same as we've always done, but just doing vastly more. But none the less, loads of consultants will write reports about how it will save you money and lots of people will be disappointed. >> I want to poke at that for a second. (laughing) I don't disagree with Javons Paradox when it comes to power, but example, say you know, our host for this event, A Cloud Guru. >> Yeah. >> They're priced to deliver per user is way lower than if they'd have done this the traditional way and I've heard many examples here at the show already where they've said, oh if I had built it this way, you know, it's now an order of magnitude less dollars, so. >> Let's forget order of mag, let's go many orders of magnitude. So from now to say the 1980s, for a thousand dollars, I can get a million times more compute resource than I could back then. Has my IT budget reduced a million fold during that time? And the answer is >> Yeah. >> What, my IT budget has reduced a million fold? >> No, no, no my IT budget has not reduced a million fold. >> Not at all, because we've just ended up doing vastly more stuff. >> Yeah, yeah. >> So the point is, yes. >> Budgets are always flat, yes. >> So the point is yes, we will be able to do the same things but more efficiently, but your IT budget doesn't reduce because we end up doing more things. So we're in competition, say, you and me and say you evolve, you use these environments you don't reduce your IT spending, you do more things, I'm now having to spend more and more just to try and keep up with you. So eventually I'm forced to adopt to that new world. So what happens is the individual acts become more efficient, but because we do more, we don't save anything. >> You know, want to look at kind of, maps versus strategy. >> Okay. >> I guess one of the things, if I'm talking to the typical Enterprise CIO or Board and they say, oh, well, a year ago I heard about Serverless, or today I heard about Serverless, you know, the strategy is going to change greatly because this is changing so rapidly, how do you help companies understand when things are changing so fast, how do I set a strategy for today? How long do I keep it? How often do I revisit it? >> So, if you map an environment, like all maps, they're dynamic, so you're constantly adapting and changing them as the environment is changing. So, when you look at, you have the purpose of your company, you have the landscape you're operating in, there are a number of climatic patents, about 30 of them, which impact that environment, will change it, so you need to understand those. Then there's sort of university useful patents known as doctrine, then there's game play. Now, for most organizations, because they cannot see the environment, they cannot distinguish, or may just be completely oblivious to any of this, so when they were talking about change, if I look at how things evolve from genesis, custom built product commodity, most organizations will go genesis, that's an innovation, every custom built feature differentiation of a product's an innovation, every shift in product to utility is an innovation, so all they see is innovation, innovation, innovation. And therefore, it's very easy to get sucked in to one size fits all methods work. One size innovation programs, where in fact, the genesis you would be using something like a lightweight XP, the product development, much more lean enterprise, so SCRUM and MVP and the utility is much more outsourcing or Six Sigma. So you should be using multiple techniques and multiple methods and most organizations aren't in that position. And if they're not in that position, of being able to see the environment, it's difficult to see where to attack, it's difficult to understand why here over there, it's difficult to manipulate the market. So, what happens is most organizations work on gut feel, whatever's popular in HPR and just act. And you can call that strategy if you wish. >> Alright, so I wish we could talk for another couple of hours, but want to give you the final take away >> Yes. >> Serverless today, how should people be thinking about it and what should they be looking for over the next six to 12 months in this space? >> So, the key thing about Serverless is we're seeing a shift from platform from product to utility, so you should be developing skills in that space. And we're seeing co-evolution of practice. By that, we mean there is a new set of practices combining finance and development together. What those practices are, we don't know yet. You have to experiment and explore. That's why attending events and being involved in building stuff will help you discover those practices. So today if your company, well it depends on your position, so if you're a company which is behind the game, you, say, haven't gone into infractructure as a service, you're not doing DevOps, you're own people are resistant to this change cause the other vendors say you're going to lose their jobs and blah, then rather then embarking on a five to seven year program, 'cause that's how long it will take to do that, you should move up the stack and start with Serverless and learning those practices. 'Cause no one knows them well, so you can take your people who've got inertia and re-train them in that space overcoming that inertia and give yourself a path forward. So, depends on your position, but I think most companies should be experimenting in this space. >> Alright, well Simon Wardley, it's a pleasure to catch up with you today. >> Delight. >> Hope to have you back on theCUBE at another event soon. Thank you so much for watching theCUBE.

>> Announcer: From Hell's Kitchen in New York City, it's the Cube on the ground at Serverlessconf Brought to you by SiliconAngle Media. >> Hi, I'm Stu Miniman with the Cube, here at Serverlessconf in Hell's Kitchen, New York City. Our first time doing the Cube here. Happy to welcome back to the program, a multi-time guest, Mark Nunnikhoven who is the Vice President of Cloud Research at Trend Micro. Mark, great to see you. >> Thanks for having me. Great to see you Stu. >> Alright, so Mark repeat after me. >> Stu and Mark: Security is everybody's responsibility. >> Yeah. >> So you did a keynote talking about security, and I love, unfortunately I didn't get to see it in person, but I feel like I was there 'cause we had the Twitter's and the commentary. >> Yeah. >> And stuff like that. So security, it's a non-issue right? Serverless it's all set, containers and everything before it, everything's secure right? >> Yeah. As you know from looking at the headlines, we do security really well in the IT community. So you can sleep well at night. We don't have to worry about anything. No, unfortunately it continues to be a challenge, and the point of the keynote yesterday was, sort of, give the state of the nation, how we're doing in the Serverless environment. And the good news is we're doing well in security for Serverless designs, but the bad news is not through any individual or purpose action. Simply by just building in these methods, we get a huge amount of security advantages. >> Yeah. >> But we can do better. >> Alright so Mark, what can we learn? It's funny, we see these repetitive things go on in the industry. It's like "Oh well, I'm just going to use Sass." "I don't need to worry about security, right?" "Oh, I'm going to go public Cloud, they'll take care of it for me." >> Yeah. >> Now containers, Serverless it feels like we have the same trope over, and over, and over again, right? >> We do, very very much so. And one of the things I called out yesterday was actually highlighting how the OWASP Top 10, which is the 10 most common vulnerabilities in web applications, have not really changed since 2010. Yet we didn't have even the concept of Serverless in 2010, but we're still making the same mistakes. SQL injection, still the top mistake that we've been making for the last decade. >> Alright, so we're talking about security. Let's step back for a second. So I believe a lot of the people watching these interviews are going to be like "Serverless, I don't get it." I love the, the Cloud Guru folks have the t-shirt, the update of the Cloud one. There is no Cloud, it's just somebody else owns the computer now. I forget the full thing. >> Somebody else's informal execution environment that last's for milliseconds, something along that. >> So what from your standpoint, you've been talking to a lot of customers >> Yeah. >> that you're speaking at this conference. You know, the what and the why of Serverless? >> Yeah, so Serverless is really that sort of, I won't say conclusion, but the logical next step of Cloud where you start to realize, when you move out of your own data center where you were doing everything, and you move into the Cloud and go, well half of the responsibility is on Amazon, or Google, or Microsoft, or whoever. And then you go, well hold on a second, why am I even managing Windows or Linux? What advantage is that to me? I make widgets, or I sell shirts or whatever. And so you move up into something like containers, and you ask the same question. Go, well why am I even running those? Serverless is that last step on the current line of going, I don't have to run any of this stuff. I can just write code that's directly tied to my business. >> Yeah, and I like how you said it's the next step. I think back to science, and it was like when we found the atom. Everybody was super excited, and then oh, there were protons and neutrons, and they were like oh my gosh, and electrons and everything. And then they're like "Oh and then there's the quark." >> Yeah. >> Everything like that. So the digger, the further down we deep, but what is the value of that? So we went from the server, to virtualized environments, to microservices, to containers. Why is that important? What's the business outcome that people are getting when they get excited and start playing with Serverless? >> For sure, so there's really two main points for me. One is that you have a direct tie between IT and the business, both from performance as well as cost. So now you can actually say that application had cost me $1.10 per transaction, and I normally make $9 on each transaction. So this is good, let's continue to invest there. So there's finally a breakdown between the separation, and you get that unity with the business and IT. And the second is accessibility. Because there's far less infrastructure and plumbing to worry about, you have people who aren't traditionally viewed as developers, more of the business analysts, starting to actually write solutions that are far more directly in line with what you want to do as a business. >> Alright, one of the things I liked seeing in the keynotes was can we do today and what can't we do today? So web applications, great, IOT, things like the Amazon Button, or the Amazon Alexa. >> Yeah. >> All leverage that. What are some of the cool applications that you've seen leveraging Serverless today? >> Yeah, so a lot of cool robots. A friend of mine, Ben Kehoe from iRobot gave a great talk on it. A lot of their stuff leverages that, and I'm a nerd, I love robots. >> Who doesn't like robots? >> Exactly, right? >> We welcome our robot overlords here at the Cube. >> Absolutely. And if they're listening, when they process this, thank you for your service. But yeah, there's a lot of great things where we're crossing out of the digital world into the real world. Because we can connect these things up with the advantage of Serverless. We don't have to build out a huge infrastructure. If you need smart lighting, if you need smart appliances, all of the IOT world, it's all Serverless. >> Yeah. So I'm going to bring up this word >> Yeah. >> That has some weight to it enterprise. >> Uh oh, let me brace. Yeah. >> So companies, we're talking, the Cloud is being used for whole businesses and everything like that. Is Serverless for, it's web, and robots, and cool toys, and everything like this. What are you seeing? What are the limitations, and does this become a predominant operating model in the future? >> Yeah, there's a lot of hesitancy in the enterprise because they're not familiar with it. >> Yeah. >> But realistically, any enterprise today should have a very simple, sort of, fall down model. When they're building something new, start at Serverless. If that doesn't meet your needs, put it in a container. If that doesn't meet your needs, build a server. Again, you want to do less work. The challenge, again, is comfort level. Serverless breaks a lot of our tooling. >> Yeah. >> So you need to learn a lot of stuff, but it's definitely where enterprises should be looking today if they want to get ahead. >> Okay, and Mark what advice do you give to companies today as they think about security across some of these various environments? >> Well you led the cheer at the start. Security is everybody's responsibility. From a security practitioners side, point of view, we've done ourselves a disservice in isolating ourselves in teams and not talking to people. We need to be educators within our organizations to help people understand what they can do. It goes all the way back to the Mythical Man-Month. It's easier to squash a bug before you ever write it, rather than when it's deployed to millions of people. Same thing for security, the earlier you're on it, the more people are looking at it, the better off you're going to be. >> Alright Mark, I want to give you the final word. Take aways, the event isn't done, but for people that aren't familiar where do they get started? Where should they dig in for Serverless? >> Yeah, there's a ton of great content here. So this is the fifth Serverless event. A lot of the old talks are up on YouTube, and Cloud Guru's done a fantastic job on pulling this community together. Check out all that stuff. The major providers, all of them are here. All of them have excellent entry level projects to help you get rolling and really that's the best way to start. Fire up the console, start building something. Why not? >> Alright Mark, really appreciate you joining. Thank for sharing with the community here, our community. Look forward to seeing you at many more events, and thank you so much for watching the Cube. (upbeat music)

>> Announcer: Live from Midtown Manhattan, it's theCUBE, covering Big Data New York City 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsor. >> Okay, welcome back, everyone, to our live special CUBE coverage in New York City in Manhattan, we're here in Hell's Kitchen for theCUBE's exclusive coverage of our Big Data NYC event and Strata Data, which used to be called Strata Hadoop, used to be Hadoop World, but our event, Big Data NYC, is our fifth year where we gather every year to see what's going on in big data world and also produce all of our great research. I'm John Furrier, the co-host of theCUBE, with Peter Burris, head of research. Our next guest, Itamar Ankorion, who's the Chief Marketing Officer at Attunity. Welcome back to theCUBE, good to see you. >> Thank you very much. It's good to be back. >> We've been covering Attunity for many, many years. We've had many conversations, you guys have had great success in big data, so congratulations on that. But the world is changing, and we're seeing data integration, we've been calling this for multiple years, that's not going away, people need to integrate more. But with cloud, there's been a real focus on accelerating the scale component with an emphasis on ease of use, data sovereignty, data governance, so all these things are coming together, the cloud has amplified. What's going on in the big data world, and it's like, listen, get movin' or you're out of business has pretty much been the mandate we've been seeing. A lot of people have been reacting. What's your response at Attunity these days because you have successful piece parts with your product offering? What's the big update for you guys with respect to this big growth area? >> Thank you. First of all, the cloud data lakes have been a major force, changing the data landscape and data management landscape for enterprises. For the past few years, I've been working closely with some of the world's leading organizations across different industries as they deploy the first and then the second and third iteration of the data lake and big data architectures. And one of the things, of course, we're all seeing is the move to cloud, whether we're seeing enterprises move completely to the cloud, kind of move the data lakes, that's where they build them, or actually have a hybrid environment where part of the data lake and data works analytics environment is on prem and part of it is in the cloud. The other thing we're seeing is that the enterprises are starting to mix more of the traditional data lake, the cloud is the platform, and streaming technologies is the way to enable all the modern data analytics that they need, and that's what we have been focusing on on enabling them to use data across all these different technologies where and when they need it. >> So, the sum of the parts is worth more if it's integrated together seems to be the positioning, which is great, it's what customers want, make it easier. What is the hard news that you guys have, 'cause you have some big news? Let's get to the news real quick. >> Thank you very much. We did, today, we have announced, we're very excited about it, we have announced a new big release of our data integration platform. Our modern platform brings together Attunity Replicate, Attunity Compose for Hive, and Attunity Enterprise Manager, or AEM. These are products that we've evolved significantly, invested a lot over the last few years to enable organizations to use data, make data available, and available in the real time across all these different platforms, and then, turn this data to be ready for analytics, especially in Hive and Hadoop environments on prem and now also in the cloud. Today, we've announced a major release with a lot of enhancements across the entire product line. >> Some people might know you guys for the Replicate piece. I know that this announcement was 6.0, but as you guys have the other piece part to this, really it's about modernization of kind of old-school techniques. That's really been the driver of your success. What specifically in this announcement makes it, you know, really work well for people who move in real time, they want to have good data access. What's the big aha for the customers out there with Attunity on this announcement? >> That's a great question, thank you. First of all is that we're bringing it all together. As you mentioned, over the past few years, Attunity Replicate has emerged as the choice of many Fortune 100 and other companies who are building modern architectures and moving data across different platforms, to the cloud, to their lakes, and they're doing it in a very efficient way. One of the things we've seen is that they needed the flexibility to adapt as they go through their journey, to adapt different platforms, and what we give them with Replicate was the flexibility to do so. We give them the flexibility, we give them the performance to get the data and efficiency to move only the change of the data as they happen and to do that in a real-time fashion. Now, that's all great, but once the data gets to the data lake, how do you then turn it into valuable information? That's when we introduced Compose for Hive, which we talked about in our last session a few month ago, which basically takes the next stage in the pipeline picking up incremental, continuous data that is fed into the data lake and turning those into operational data store, historical data stores, data store that's basically ready for analytics. What we've done with this release that we're really excited about is putting all of these together in a more integrated fashion, putting Attunity Enterprise Manager on top of it to help manage larger scale environments so customers can move faster in deploying these solutions. >> As you think about the role that Attunity's going to play over time, though, it's going to end up being part of a broader solution for how you handle your data. Imagine for a second the patterns that your customers are deploying. What is Attunity typically being deployed with? >> That's a great question. First of all, we're definitely part of a large ecosystem for building the new data architecture, new data management with data integration being more than ever a key part of that bigger ecosystem because as all they actually have today is more islands with more places where the data needs to go, and to your point, more patterns in which the data moves. One of those patterns that we've seen significantly increase in demand and deployment is streaming. Where data used to be batch, now we're all talking about streaming. Kafka has emerged as a very common platform, but not only Kafka. If you're on Amazon Web Services, you're using Kinesis. If you're in Azure, you're using Azure Event Hubs. You have different streaming technologies. That's part of how this has evolved. >> How is that challenge? 'Cause you just bring up a good point. I mean, with the big trend that customers want is they want either the same code basis on prem and that they have the hybrid, which means the gateway, if you will, to the public cloud. They want to have the same code base, or move workloads between different clouds, multi-cloud, it seems to be the Holy Grail, we've identified it. We are taking the position that we think multi-cloud will be the preferred architecture going forward. Not necessarily this year, but it's going to get there. But as a customer, I don't want to have to rebuild employees and get skill development and retraining on Amazon, Azure, Google. I mean, each one has its own different path, you mentioned it. How do you talk to customers about that because they might be like, whoa, I want it, but how do I work in that environment? You guys have a solution for that? >> We do, and in fact, one of the things we've seen, to your point, we've seen the adoption of multiple clouds, and even if that adoption is staged, what we're seeing is more and more customers that are actually referring to the term lock-in in respect to the cloud. Do we put all the eggs in one cloud, or do we allow ourselves the flexibility to move around and use different clouds, and also mitigate our risk in that respect? What we've done from that perspective is first of all, when you use the Attunity platform, we take away all the development complexity. In the Attunity platform, it is very easy to set up. Your data flow is your data pipelines, and it's all common and consistent. Whether you're working on prem, whether you work on Amazon Web Services, on Azure, or on Google or other platforms, it all looks and feels the same. First of all, and you solve the issue of the diversity, but also the complexity, because what we've done is, this is one of the big things that Attunity is focused on was reducing the complexity, allowing to configure these data pipelines without development efforts and resources. >> One of the challenges, or one of the things you typically do to take complexity out is you do a better job of design up front. And I know that Attunity's got a tool set that starts to address some of of these things. Take us a little bit through how your customers are starting to think in terms of designing flows as opposed to just cobbling together things in a bespoke way. How is that starting to change as customers gain experience with large data sets, the ability, the need to aggregate them, the ability to present them to developers in different ways? >> That's a great point, and again, one of the things we've focused on is to make the process of developing or configuring these different data flows easy and modular. First, while in Attunity you can set up different flows in different patterns, and you can then make them available to others for consumption. Some create the data ingestion, or some create the data ingestion and then create a data transformation with Compose for Hive, and with Attunity Enterprise Manager, we've now also introduced APIs that allow you to create your own microservices, consuming and using the services enabled by the platform, so we provide more flexibility to put all these different solutions together. >> What's the biggest thing that you see from a customer standpoint, from a problem that you solve? If you had to kind of lay it out, you know the classic, hey, what problem do you solve? 'Cause there are many, so take us through the key problem, and then, if there's any secondary issues that you guys can address customers, that seems the way conversation starts. What are key problems that you solve? >> I think one of the major problems that we solve is scale. Our customers that are deploying data lakes are trying to deploy and use data that is coming, not from five or 10 or even 50 data sources, we work at hundreds going on thousands of data sources now. That in itself represents a major challenge to our customers, and we're addressing it by dramatically simplifying and making the process of setting those up very repeatable, very easy, and then providing the management facility because when you have hundreds or thousands, management becomes a bigger issue to operationalize it. We invested a lot in a management facility for those, from a monitoring, control, security, how do you secure it? The data lake is used by many different groups, so how do we allow each group to see and work only on what belongs to that group? That's part it, too. So again, the scale is the major thing there. The other one is real timeliness. We talked about the move to streaming, and a lot of it is in order to enable streaming analytics, real-time analytics. That's only as good as your data, so you need to capture data in real time. And that of course has been our claim to fame for a long time, being the leading independent provider of CDC, change data capture technology. What we've done now, and also expanded significantly with the new release, version six, is creating universal database streaming. >> What is that? >> We take databases, we take databases, all the enterprise databases, and we turn them into live streams. When you think, by the way, by the most common way that people have used, customers have used to bring data into the lake from a database, it was Scoop. And Scoop is a great, easy software to use from an open source perspective, but it's scripting and batch. So, you're building your new modern architecture with the two are effectively scripting and batch. What we do with CDC is we enable to take a database, and instead of the database being something you come to periodically to read it, we actually turn it into a live feed, so as the data changes in the database, we stream it, we make it available across all these different platforms. >> Changes the definition of what live streaming is. We're live streaming theCUBE, we're data. We're data streaming, and you get great data. So, here's the question for you. This is a good topic, I love this topic. Pete and I talk about this all the time, and it's been addressed in the big data world, but it's kind of, you can see the pattern going mainstream in society globally, geopolitically and also in society. Batch processing and data in motion are real time. Streaming brings up this use case to the end customer, which is this is the way they've done it before, certainly store things in data lakes, that's not going to go away, you're going to store stuff, but the real gain is in motion. >> Itamar: Correct. >> How do you describe that to a customer when you go out and say, hey, you know, you've been living in a batch world, but wake up to the real world called real time. How do you get to them to align with it? Some people get it right away, I see that, some people don't. How do you talk about that because that seems to be a real cultural thing going on right now, or operational readiness from the customer standpoint? Can you just talk through your feeling on that? >> First of all, this often gets lost in translation, and we see quite a few companies and even IT departments that when you talk, when they refer to real time, or their business tells them we need real time, what they understand from it is when you ask for the data, the response will be immediate. You get real time access to the data, but the data is from last week. So, we get real time access, but for last week's data. And that's what we try to do is to basically say, wait a second, when you mean real time, what does real time mean? And we start to understand what is the meaning of using last week's data versus, or yesterday's data, over the real time data, and that makes a big difference. We actually see that today the access, the availability, the availability to act on the real time data, that's the frontier of competitive differentiation. That's what makes a customer experience better, that's what makes the business more operationally efficient than the competition. >> It's the data, not so much the process of what they used to do. They're version of real time is I responded to you pretty quickly. >> Exactly, the other thing that's interesting is because we see it with, again, change of the capture becoming a critical component of the modern data architecture. Traditionally, we used to talk about different type of tools and technology, now CDC itself is becoming a critical part of it, and the reason is that it serves and it answers a lot of fundamental needs that are now becoming critical. One is the need for real-time data. The other one is efficiency. If you're moving to the cloud, and we talked about this earlier, if you're data lake is going to be in the cloud, there's no way you're going to reload all your data because the bandwidth is going to get in the way. So, you have to move only the delta. You need the ability to capture and move only the delta, so CDC becomes fundamental both in enabling the real time as well the efficient, the low-impact data integration. >> You guys have a lot of partners, technology partners, global SIs, resellers, a bunch of different partnership levels. The question I have for you, love to get your reaction and share your insight into is, okay, as the relationship to the customer who has the problem, what's in it for me? I want to move my business forward, I want to do digital business, I need to get up my real-time data as it's happening. Whether it's near real time or real time, that's evolution, but ultimately, they have to move their developers down a certain path. They'll usually hire a partner. The relationship between partners and you, the supplier to the customer, has changed recently. >> That's correct. >> How is that evolving? >> First of all, it's evolving in several ways. We've invested on our part to make sure that we're building Attunity as a leading vendor in the ecosystem of they system integration consulting companies. We work with pretty much all the major global system integrators as well as regional ones, boutique ones, that focus on the emerging technologies as well as get the modern analytic-type platforms. We work a lot with plenty of them on major corporate data center-level migrations to the cloud. So again, the motivations are different, but we invest-- >> More specialized, are you seeing more specialty, what's the trend? >> We've been a technology partner of choice to both Amazon and Microsoft for enabling, facilitating the data migration to the cloud. They of course, their select or preferred group of partners they work with, so we all come together to create these solutions. >> Itamar, what's the goals for Attunity as we wrap up here? I give you the last word, as you guys have this big announcement, you're bringing it all together. Integrating is key, it's always been your ethos in the company. Where is this next level, what's the next milestone for you guys? What do you guys see going forward? >> First of all, we're going to continue to modernize. We're really excited about the new announcement we did today, Replicate six, AEM six, a new version of Compose for Hive that now also supports small data lakes, Aldermore, Scaldera, EMR, and a key point for us was expanding AEM to also enable analytics on the data we generate as data flows through it. The whole point is modernizing data integration, providing more intelligence in the process, reducing the complexity, and facilitating the automation end-to-end. We're going to continue to solve, >> Automation big, big time. >> Automation is a big thing for us, and the point is, you need to scale. In order to scale, we want to generate things for you so you don't to develop for every piece. We automate the automation, okay. The whole point is to deliver the solution faster, and the way we're going to do it is to continue to enhance each one of the products in its own space, if it's replication across systems, Compose for Hive for transformations in pipeline automation, and AEM for management, but also to create integration between them. Again, for us it's to create a platform that for our customers they get more than the sum of the parts, they get the unique capabilities that we bring together in this platform. >> Itamar, thanks for coming onto theCUBE, appreciate it, congratulations to Attunity. And you guys bringing it all together, congratulations. >> Thank you very much. >> This theCUBE live coverage, bringing it down here to New York City, Manhattan. I'm John Furrier, Peter Burris. Be right back with more after this short break. (upbeat electronic music)

>> Announcer: Live, from Midtown Manhattan, it's theCUBE. Covering Big Data, New York City, 2017. Brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Okay welcome back, everyone. Live in Manhattan, New York City, in Hell's Kitchen, this is theCUBE's special event, our annual CUBE-Wikibon Research Big Data event in Manhattan. Alongside Strata, Hadoop; formerly Hadoop World, now called Strata Data, as the world continues. This is our annual event; it's our fifth year here, sixth overall, wanted to kind of move from uptown. I'm John Furrier, the co-host of theCUBE, with Peter Burris, Head of Research at SiliconANGLE and GM of Wikibon Research. Our next guest is Tim Smith, who's the SVP of technical operations at AppNexus; technical operations for large scale is an understatement. But before we get going; Tim, just talk about what AppNexus as a company, what you guys do, what's the core business? >> Sure, AppNexus is the second largest digital advertising marketplace after google. We're an internet technology company that harnessed, we harness data and machine learning to power the companies that comprise the open internet. We began by building a powerful technology platform, in which we embedded core capabilities, tools and features. With me so far? >> Yeah, we got it. >> Okay, on top of that platform, we built a core suite of cloud-based enterprise products that enable the buying and selling of digital advertising, and a scale-transparent and low-cost marketplace where other companies can transact; either using our enterprise products, or those offered by other companies. If you want to hear a little about the daily peaks, peak feeds and speeds, it is Strata, we should probably talk about that. We do about 11.8 billion impressions transacted on a daily basis. Each of those is a real-time auction conducted in a fraction of a second, well under half a second. We see about 225 billion impressions per day, and we handle about 5 million queries per second at peak load. We produce about 150 terabytes of data each day, and we move about 400 gigabits into and out of the internet at peak, all those numbers are daily peaks. Makes sense? >> Yep. >> Okay, so by way of comparison, which might be useful for people, I believe the NYSE currently does roughly 2 million trades per day. So if we round that up to 3 million trades a day and assume the NYSE were to conduct that volume every single day of the year; 7 days a week, 365 days a year, that'd be about a billion trades a year. Similarly, I believe Visa did about 28-and-a-half billion transactions in their fiscal third quarter. I'll round that up to 30 billion, and average it out to about 333 million transactions per day and annualize it to about 4 billion transactions per year. Little bit of math, but as I mentioned, AppNexus does an excess of 10 billion transactions per day. And so it seems reasonable to say that AppNexus does roughly 10 times the transaction volume in one day, than the NYSE does in a year. And similarly, it seems reasonable to say that AppNexus daily does more than two times the transaction volume that Visa does in a year. Obviously, these are all just very rough numbers based on publicly available information about the NYSE and Visa, and both the NYSE and Visa do far, far more volume than AppNexus when measured in terms of dollars. So given our volumes, it's imperative that AppNexus does each transaction with the maximum efficiency and lowest reasonable possible cost, and that is one of the most challenging aspects of my job. >> So thanks for spending the time to give the overview. There's a lot of data; I mean 10 billion a day is massive volume. I mean the internet, and you see the scale, is insane. We're in a new era right now of web-scale. We've seen it in Facebook, and it's enormous. It's only going to get bigger, right? So on the online ad tech, you guys are essentially doing like a Google model, that's not everything but Google, which is still huge numbers. Then you include Microsoft and everybody else. Really heavy lifting, IT-like situation. What's the environment like? And just talk about, you know, what's it like for you guys. Because you got a lot of opp's, I mean terms of dev opp's. You can't break anything, because that 10 billion transaction or near, it's a significant impact. So you have to have everything buttoned-up super tight, yet you got to innovate and grow with the future growth. What's the IT environment like? >> It's interesting. We have about 8,000 servers spread across about seven data centers on three continents, and we run, as you mentioned, around the clock. There's no closing bell; downtime is not acceptable. So when you look at our environment, you're talking about four major categories of server complexes. We have real-time processing, which is the actual ad serving. We have a data pipeline, which is what we call our big data environment. We also have client-facing environment and an infrastructure environment. So we use a lot of different tools and applications, but I think the most relevant ones to this discussion are Hadoop and its friends HDFS, and Hive and Spark. And then we use the Vertica Analytics Platform. And together Hadoop and its friends, and Vertica comprise our entire data pipeline. They're both very disk-intensive. They're cluster based applications, and it's a lot of challenge to keep them up and running. >> So what are some of those challenges? Just explain a little bit, because you also have a lot of opportunity. I mean, it's money flowing through the air, basically; digital air, if you will. I mean, they got a lot of stuff happening. Take us through the challenges. >> You know, our biggest apps are all clustered. And all of our clusters are built with commodity servers, just like a lot of other environments. The big data app clusters traditionally have had internal disks, while almost all of our other servers are very light on disk. One of the biggest challenges is, since the server is the fundamental building block of a cluster, then regardless of whether you need more compute or more storage, you always have to add more servers to get it. That really limits flexibility and creates a lot of inefficiencies, and I really, really am obsessive about reducing and eliminating inefficiencies. So, with me so far? >> Yep. >> Great. The inefficiencies result from two major factors. First, not all workloads require the same ratio of compute to storage. Some workloads are more compute-intensive, and others are really less dependent on storage, while other workloads require a lot more storage. So we have to use standard server configurations and as a result, we wind up with underutilized compute and storage. This is undesirable, it's inefficient, yet given our scale, we have to use standardized configurations. So that's the first big challenge. The second is the compute to disk ratio. It's generally fixed when you buy the servers. Yes, we can certainly add more disks in the field, but that's a labor intensive, and it's complicated from a logistics and an asset management standpoint, and you're fundamentally limited by the number of disk slots in the server. So now you're right back into the trap of more storage requires more servers, regardless of whether you need more compute or not. And then you compound the inefficiencies. >> Couldn't you just move the resources from, unused resources, from one cluster to the other? >> I've been asked that a lot; and no, it's just not that simple. Each application cluster becomes a silo due to its configuration of storage and compute. This means you just can't move servers from clusters because the clusters are optimized for the workloads, and the fact that you can't move resources from one cluster to another, it's more inefficiencies. And then they're compounded over time since workloads change, and the ideal ratio of compute-to-storage changes. And the end result is unused resources trapped in silos and configurations that are no longer optimized for your workload. And there's only really one solution that we've been able to find. And to paraphrase an orator far, far more talented than I am, namely Ronald Reagan, we need to open this gate, tear down these silos. The silos just have to go away. They fundamentally limit flexibility and efficiency. >> What were some of the other issues caused by using servers with internal drives? >> You have more maintenance, you've got to deal with the logistics. But the biggest problem is service and storage have significantly different life cycles. Servers typically have a three year life cycle before they're obsolete. Storage typically is four to six years. You can sometimes stretch that a little further with the storage. Inside the servers that are replaced every 3 years, we end up replacing storage before the end of its effective lifetime; that's inefficient. Further, since the storage is inside the servers, we have to do massive data migrations when we replace servers. Migrations, they're time consuming, they're logistically difficult, and they're high risk. >> So how did DriveScale help you guys? Because you guys certainly have a challenging environment, you laid out the the story, and we appreciate that. How did DriveScale help you with the challenges? >> Well, what we really wanted to do was disaggregate storage from servers, and DriveScale enables us to do that. Disaggregating resources is a new term in the industry, but I think lot of people are focusing on it. I can explain it if you think that would make sense. >> What do you mean by disaggregating resources? Can you explain that, and how it works? >> Sure, so instead of buying servers with internal drives, we now buy diskless servers with JBODs. And DriveScale lets us easily compose servers with whatever amount of disk storage we need, from the server resource pool and the disk resource pool; and they're separate pools. This means we have the right balance of compute and storage for each workload, and we can easily adjust it over time. And all of this is done via software, so it's easy to do with a GUI or in our case, at our scale, scripting. And it's done on demand, and it's much more efficient. >> How does it help you with the underutilized resource challenge you mentioned earlier? >> Well, since we can add and remove resources from each cluster, we can manage exactly how much compute power and storage is deployed for each workload. Since this is all done via software, it can be done quickly and easily. We don't have to send a technician into a data center to physically swap drives, add drives, move drives. It's all done via software and it's very, very efficient. >> Can you move resources between silos? >> Well, yes and no. First off, our goal is no more silos. That said, we still have clusters, and once we completely migrate to DriveScale, all of our compute and storage resources will be consolidated into just a few common pools. And disk storage will no longer differentiate pools; thus, we have fewer pools. For more, we have fewer pools and can use the resources in each pool for more workloads. And when our needs change and they always do, we can reallocate resources as needed. >> What of the life cycle management challenge? How you guys address that? >> Well that's addressed with DriveScale. The compute and the storage are now disaggregated or separated into diskless servers and JBODs, so we can upgrade one without touching the other. We want to upgrade servers to take advantage of new processors or new memory architectures, we just replace the servers, re-combine the disks with the new servers, and we're back up and operating. It saves the cost of buying new disks when we don't need to, and it also simplifies logistics and reduces risk, as we no longer have to run the old plant and the new plant concurrently, and do a complicated data migration. >> What about this qualifying server and storage vendors? Do you still do that? Or how's that impact -- >> We actually don't have to do it. We're still using the same server vendor. We've used Dell for many, many years, we continue to use them. We are using them for storage and there was no real work, we just had to add DriveScale into the mix. >> What's it like working with DriveScale? >> They're really wonderful to work with. They have a really seasoned team. They were at Sun Microsystems and Cisco, they built some of the really foundational products that changed the internet, that the internet was built on. They're really talented, they really bright, and they're really focused on customer success. >> Great story, thanks for sharing that. My final question for you is, you guys have a very big, awesome environment, you've got a lot of scale there. It's great for a startup to get into an environment like this, because one, they could get access to the data, work with a good team like you have. What's it like working with a startup? >> You know it's always challenging at first; too many things to do. >> They got talented guys. Most of the startups, those early day startups, they got all their A players out there. >> They have their A players, and we've been very pleased working with them. We're dealing with the top talent, some of the top talent in the industry, that created the industry. They have a proven track record. We really don't have any concerns, we know they're committed to our success and they have a great team, and great investors. >> A final, final question. For your friends out there are watching, and other practitioners who are trying to run things at scale with a cloud. What's your advice to them? You've been operating at scale, and a lot of, billions of transactions, I mean huge; it's only going to get bigger. Put your IT friendly advice hat on. What's the mindset of operators out there, technical op's, as dev op's comes in seeing a lot of that. What do people need to be thinking about to run at scale? >> There's no magic silver bullet. There's no magic answers. The public cloud is very helpful in a lot of ways, but you really have to think hard about your economics, you have to think about your scale. You just have to be sure that you're going into each decision knowing that you've looked at the costs and the benefits, the performance, the risks, and you don't expect there to be simple answers. >> Yeah, there's no magic beans as they say. You've got to make it work for the business. >> No magic beans, I wish there were. >> Tim, thanks so much for the story. Appreciate the commentaries. Live coverage at Big Data NYC, it's theCUBE. Be back with more after this short break. (upbeat techno music)

>> Announcer: Live from midtown Manhattan, it's theCUBE, covering BigData New York City 2017, brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Okay welcome back everyone here live in New York City. This is theCUBE special presentation of our annual event with theCUBE and Wikibon Research called BigData NYC, it's our own event that we have every year, celebrating what's going on in the big data world now. It's evolving to all data, cloud applications, AI, you name it, it's happening. In the enterprise, the impact is huge for developers, the impact is huge. I'm John Furrier, cohost of the theCUBE, with Peter Burris, Head of Research, SiliconANGLE Media and General Manager of Wikibon Research. Our next guest is Jagane Sundar, who's the CTO of WANdisco, Cube alumni, great to see you again as usual here on theCUBE. >> Thank you John, thank you Peter, it's great to be back on theCUBE. >> So we've been talking the big data for many years, certainly with you guys, and it's been a great evolution. I don't want to get into the whole backstory and history, we covered that before, but right now is a really, really important time, we see you know the hurricanes come through, we see the floods in Texas, we've seen Florida, and Puerto Rico now on the main conversation. You're seeing it, you're seeing disasters happen. Disaster recovery's been the low hanging fruit for you guys, and we talked about this when New York City got flooded years and years ago. This is a huge issue for IT, because they have to have disaster recovery. But now it's moving more beyond just disaster recovery. It's cloud. What's the update from WANdisco? You guys have a unique perspective on this. >> Yes, absolutely. So we have capabilities to replicate between the cloud and Hadoop multi data centers across geos, so disasters are not a problem for us. And we have some unique technologies we use. One of the things we do is we can replicate in an active-active mode between different cloud vendors, between cloud and on-prem Hadoop, and we are the only game in town. Nobody else can do that. >> So okay let me just stop right there. When you say the only game in town I got a little skeptic here. Are you saying that nobody does active-active replication at all? >> That is exactly what I'm saying. We had some wonderful announcements from Hortonworks, they have a great product called the Dataplane. But if you dig deep, you'll find that it's actually an active-passive architecture, because to do active-active, you need this capability called the Paxos algorithm for resolving conflict. That's a very hard algorithm to implement. We have over 10 years' experience in that. That's what gives us our ability to do this active-active replication, between clouds, between on-prem and cloud. >> All right so just to take that a step further, I know we're having a CTO conversation, but the classic cliche is skate to where the puck is going to be. So you kind of didn't just decide one morning you're going to be the active-active for cloud. You kind of backed into this. You know the world spun in your direction, the puck came to you guys. Is that a fair statement? >> That is a very fair statement. We've always known there's tremendous value in this technology we own, and with the global infrastructure trends, we knew that this was coming. It wasn't called the cloud when we started out, but that's exactly what it is now, and we're benefiting from it. >> And the cloud is just a data center, it's just, you don't own it. (mumbles) Peter, what's your reaction to this? Because when he says only game in town, implies some scarcity. >> Well, WANdisco has a patent, and it actually is very interesting technology, if I can summarize very quickly. You do continuous replication based on writes that are performed against the database, so that you can have two writers and two separate databases and you guarantee that they will be synchronized at some point in time because you guarantee that the writing of the logs and the messaging to both locations >> Absolutely. >> in order, which is a big issue. You guys put a stamp on the stuff, and it actually writes to the different locations with order guaranteed, and that's not the way most replication software works. >> Yes, that's exactly right. That's very hard to do, and that's the only way for you to allow your clients in different data centers to write to the same data store, whether it's a database, a Hadoop folder, whether it's a bucket in a cloud object store, it doesn't matter. The core fact remains, the Paxos algorithm is the only way for you to do active-active replication, and ours is the only Paxos implementation that can work over the >> John: And that's patented by you guys? >> Yes, it's patented. >> And so someone to replicate that, they'd have to essentially reverse engineer and have a little twist on it to not get around the patents. Are you licensing the technology or are you guys hoarding it for yourselves? >> We have different ways of engaging with partners. We are very reasonable with that, and we work with several powerful partners >> So you partner with the technology. >> Yes. >> But the key thing, John, in answer to your question is that it's unassailable. I mean there's no argument, that is, companies move more towards a digital way of doing things, largely driven by what customers want, your data becomes more of an asset. As you data becomes more of an asset, you make money by using that data in more places, more applications and more times. That is possible with data, but the problem you end up with consistency issues, and for certain applications, it's not an issue, you're basically writing, or if you're basically reading data it's not an issue. But the minute that you're trying to write on behalf of a particular business event or a particular value proposition, then now you have a challenge, you are limited in how you can do it unless you have this kind of a technology. And so this notion of continuous replication in a world that's going to become increasingly dependent upon data, data that is increasingly distributed, data that you want to ensure has common governance and policy in place, technologies like WANdisco provides are going to be increasingly important to the overall way that a business organizes itself, institutes its work and makes sure it takes care of its data assets. >> Okay, so my next question then, thanks for the clarification, it's good input there and thanks for summarizing it like that, 'cause I couldn't have done that. But when we last talked, I always was enamored by the fact that you guys have the data center replication thing down. I always saw that as a great thing for you guys. Okay, I get that, that's an on-premise situation, you have active-active, good for disaster recovery, lot of use cases, people should be beating down your door 'cause you have a better mousetrap, I get that. Now how does that translate to the cloud? So take me through why the cloud now fits nicely with that same paradigm. >> So, I mean, these are industry trends, right. What we've found is that the cloud object stores are very, very cost effective and efficient, so customers are moving towards that. They're using their Hadoop applications but on cloud object stores. Now it's trivial for us to add plugins that enable us to replicate between a cloud object store on one side, and a Hadoop on the other side. It could also be another cloud object store from a different cloud provider on the other side. Once you have that capability, now customers are freed from lock-in from either a cloud vendor or a Hadoop vendor, and they love that, they're looking at it as another way to leverage their data assets. And we enable them to do that without fear of lock-in from any of these vendors. >> So on the cloud side, the regions have always been a big thing. So we've heard Amazon have a region down here, and there was fix it. We saw at VMworld push their VMware solution to only one western region. What's the geo landscape look like in the cloud? Does that relate to anything in your tech? >> So yes, it does relate, and one of the things that people forget is that when you create an Amazon S3 bucket, for example, you specify a region. Well, but this is the cloud, isn't it worldwide? Turns out that object store actually resides in one region, and you can use some shaky technologies like cross-region replication to eventually get the data to the other region. >> Peter: Which just boosts the prices you pay. >> Yes, not just boost the price. >> Well they're trying to save price but then they're exposed on reliability. >> Reliability, exactly. You don't know when the data's going to be there, there are no guarantees. What we offer is, take your cloud storage, but we'll guarantee that we can replicate it in a synchronous fashion to another region. Could be the same provider, could be another provider. That gives tremendous benefits to the customers. >> So you actually have a guarantee when you go to customers, say with an SLA guarantee? Do you back it up with like money back, what's the guarantee? >> So the guarantees are, you know we are willing to back it up with contracts and such like, and our customers put us through rigorous testing procedures, naturally. But we stand up to every one of those. We can scale and maintain the consistency guarantees that they need for modern businesses. >> Okay, so take me through the benefits. Who wants this? Because you can almost get kind of sucked into the complexities of it, and the nuances of cloud and everything as Peter laid out, it's pretty complex even as he simplified it. Who buys this? (laughs) I mean, who's the guy, is it the IT department, is it the ops guy, is it the facilities, who... >> So we sell to the IT departments, and they absolutely love the technology. But to go back to your initial statement, we have all these disasters happening, you know, hopefully people are all doing reasonably okay at the end of these horrible disasters, but if you're an enterprise of any size, it doesn't have to be a big enterprise, you cannot go back to your users or customers and say that because of a hurricane you cannot have access to your data. That's sometimes legally not allowed, and other times it's just suicide for a business >> And HPE in Houston, it's a huge plant down there. >> Jagane: Indeed. >> They got hit hard. >> Yep, in those sort of circumstances, you want to make sure that your data is available in multiple data centers spread throughout the world, and we give you that capability. >> Okay, what are some of the successes? Let's talk through now, obviously you've got the technology, I get that. Where's the stakes in the ground? Who's adopting it? I know you do a lot of biz dev deals. I don't know if they're actually OEM-type deals, or they're just licensing deals. Take us through to where your successes are with this technology. >> So, biz dev wise, we have a mix of OEM deals and licenses and co-selling agreements. The strong ones are all OEMs, of course. We have great partnerships with IBM, Amazon, Microsoft, just wonderful partnerships. The actual end customers, we started off selling mostly to the financial industry because they have a legal mandate, so they were the first to look into this sort of a thing. But now we've expanded into automobile companies. A lot of the auto companies are generating vast amounts of data from their cars, and you can't push all that data into a single data center, that's just not reasonable. You want to push that data into a single data store that's distributed across the world in just wherever the car is closest to. We offer that capability that nobody else can, so that we've got big auto manufacturers signed up, we've got big retailers signed up for exactly the same capability. You cannot imagine ingesting all that data into a single location. You want this replicated across, you want it available no matter what happens to any single region or a data center. So we've got tremendous success in retail, banking, and a lot of this is through partnerships again. >> Well congratulations, I got to ask, you know, what's new with you guys? Obviously you have success with the active-active. We'll dig into the Hortonworks things to check your comment around them not having it, so we'll certainly look with the Dataplane, which we like. We interviewed Rob Bearden. Love the announcement, but they don't have the active-active, we're going to document that, and get that on the record. But you guys are doing well. What's new here, what's in New York, what are some of your wins, can you just give a quick update on what's going on at WANdisco? >> Okay, so quick recap, we love the Hortonworks Dataplane as well. We think that we can build value into that ecosystem by building a plugin for them. And we love the whole technology. I have wonderful friends there as well. As for our own company, we see all of our, a lot of our business coming from cloud and hybrid environments. It's just the reality of the situation. You had, you know, 20 years ago, you had NFS, which was the great appender of all storage, but turned out to be very expensive, and you had 10 years, seven years ago you had HDFS come along, and that appended the cost model of NFS and SANs, which those industries were still working their way through. And now we have cloud object stores, which have appended the HDFS model, it's much more cost-efficient to operate using cloud object stores. So we will be there, we have replication products for that. >> John: And you're in the major clouds, you in Azure? >> Yes, we are in Azure. >> Google? >> Jagane: Yes, absolutely. >> AWS? >> AWS, of course. >> Oracle? >> Oracle, of course. >> So you got all the top four companies. >> We're in all of them. >> All right, so here's the next question is, >> And you're also in IBM stuff too. >> Yes, we're built tightly into IBM >> So you've got a pretty strong legacy >> And a monopoly. >> On the mainframe. >> Like the fiber channel of replication. (John and Jagane laugh) That was a bad analogy. I mean it's like... Well, I mean fiber channel has only limited suppliers 'cause they have unique technology, it was highly important. >> But the basic proposition is look, any customer that wants to ensure that a particular data source is going to be available in a distributed way, and you're going to have some degree of consistency, is going to look at this as an option. >> Yes. >> Well you guys certainly had a great team under your leadership, it's got great tech. The final question I have for you here is, you know, we've had many conversations about the industry, we like to pontificate, I certainly like to speculate, but now we have eight years of history now in the big data world, we look back, you know, we're doing our own event in New York City, you know, thanks to great support from you guys and other great friends in the community. Appreciate everyone out there supporting theCUBE, that's awesome. But the world's changed. So I got to ask you, you're a student of the industry, I know that and knowing you personally. What's been the success formula that keeps the winners around today, and what do people need to do going forward? 'Cause we've seen the train wreck, we've seen the dead bodies in the industry, we've kind of seen what's happened, there've been some survivors. Why did the current list of characters and companies survive, and what's the winning formula in your opinion to stay relevant as big data grows in a huge way from IoT to AI cloud and everything in between? >> I'll quote Stephen Hawking in this. Intelligence is the capability to adapt to changes. That's what keeps industries, that's what keeps companies, that what keeps executives around. If you can adapt to change, if you can see things coming, and adapt your core values, your core technology to that, you can offer customers a value proposition that's going to last a long time. >> And in a big data space, what is that adaptive key focus, what should they be focused on? >> I think at this point, it's extracting information from this volume of data, whether you use machine learning in the modern days, or whether it was simple hive queries, that's the value proposition, and making sure the data's available everywhere so you can do that processing on it, that remains the strength. >> So the whole concept of digital business suggests that increasingly we're going to see our assets rendered in some form as data. >> Yes. >> And we want to be able to ensure that that data is able to be where it needs to be when it needs to be there for any number of reasons. It's a very, very interesting world we're entering into. >> Peter, I think you have a good grasp on this, and I love the narrative of programming the world in real time. What's the phrase you use? It's real time but it's programming the world... Programming the real world. >> Yeah, programming the real world. >> That's a huge, that means something completely, it's not a tech, it's a not a speed or feed. >> Well the way we think about it, is that we look at IoT as a big information transducer, where information's in one form, and then you turn it into another form to do different kinds of work. And that big data's a crucial feature in how you take data from one form and turn it into another form so that it can perform work. But then you have to be able to turn that around and have it perform work back in the real world. There's a lot of new development, a lot of new technology that's coming on to help us do that. But any way you look at it, we're going to have to move data with some degree of consistency, we're still going to have to worry about making sure that if our policy says that that action needs to take place there, and that action needs to take place there, that it actually happens the way we want it to, and that's going to require a whole raft of new technologies. We're just at the very beginning of this. >> And active-active, things like active-active in what you're talking about really is about value creation. >> Well the thing that makes active-active interesting is, again, borrowing from your terms, it's a new term to both of us, I think, today. I like it actually. But the thing that makes it interesting is the idea that you can have a source here that is writing things, and you can have a source over there that are writing things, and as a consequence, you can nonetheless look at a distributed database and keep it consistent. >> Consistent, yeah. >> And that is a major, major challenge that's going to become increasingly a fundamental feature of our digital business as well. >> It's an enabling technology for the value creation and you call it work. >> Yeah, that's right. >> Transformation of work. Jagane, congratulations on the active-active, and WANdiscos's technology and all your deals you're doing, got all the cloud locked up. What's next? Well you going to lock up the edge? You're going to lock up the edge too, the cloud. >> We do like this notion of the edge cloud and all the intermediate steps. We think that replicating data between those systems or running consistent compute across those systems is an interesting problem for us to solve. We've got all the ingredients to solve that problem. We will be on that. >> Jagane Sundar, CTO of WANdisco, back on theCUBE, bringing it down. New tech, whole new generation of modern apps and infrastructure happening in distributed and decentralized networks. Of course theCUBE's got it covered for you, and more live coverage here in New York City for BigData NYC, our annual event, theCUBE and Wikibon here in Hell's Kitchen in Manhattan, more live coverage after this short break.

>> Announcer: Live from midtown Manhattan, it's theCUBE, covering BigData New York City 2017, brought to you by SiliconANGLE Media and its ecosystem sponsors. >> Hello everyone, welcome back to theCUBE's special BigData NYC coverage of theCUBE here in Manhattan in New York City, we're in Hell's Kitchen. I'm John Furrier, with my cohost Jim Kobielus, whose Wikibon analyst for BigData. In conjunction with Strata Data going on right around the corner, this is our annual event where we break down the big data, the AI, the cloud, all the goodness of what's going on in big data. Our next guest is Tendu Yogurtcu who's the Chief Technology Officer at Syncsort. Great to see you again, CUBE alumni, been on multiple times. Always great to have you on, get the perspective, a CTO perspective and the Syncsort update, so good to see you. >> Good seeing you John and Jim. It's a pleasure being here too. Again the pulse of big data is in New York, and it's a great week with a lot of happening. >> I always borrow the quote from Pat Gelsinger, who's the CEO of VMware, he said on theCUBE in I think 2011, before he joined VMware as CEO he was at EMC. He said if you're not out in front of that next wave, you're driftwood. And the key to being successful is to ride the waves, and the big waves are coming in now with AI, certainly big data has been rising tide for its own bubble but now the aperture of the scale of data's larger, Syncsort has been riding the wave with us, we've been having you guys on multiple times. And it was important to the mainframe in the early days, but now Syncsort just keeps on adding more and more capabilities, and you're riding the wave, the big wave, the big data wave. What's the update now with you guys, where are you guys now in context of today's emerging data landscape? >> Absolutely. As organizations progress with their modern data architectures and building the next generation analytics platforms, leveraging machine learning, leveraging cloud elasticity, we have observed that data quality and data governance have become more critical than ever. Couple of years we have been seeing this trend, I would like to create a data lake, data as a service, and enable bigger insights from the data, and this year, really every enterprise is trying to have that trusted data set created, because data lakes are turning into data swamps, as Dave Vellante refers often (John laughs) and collection of this diverse data sets, whether it's mainframe, whether it's messaging queues, whether it's relational data warehouse environments is challenging the customers, and we can take one simple use case like Customer 360, which we have been talking for decades now, right? Yet still it's a complex problem. Everybody is trying to get that trusted single view of their customers so that they can serve the customer needs in a better way, offer better solutions and products to customers, get better insights about the customer behavior, whether leveraging deep learning, machine learning, et cetera. However, in order to do that, the data has to be in a clean, trusted, valid format, and every business is going global. You have data sets coming from Asia, from Europe, from Latin America, and many different places, in different formats and it's becoming challenge. We acquired Trillium Software in December 2016, and our vision was really to bring that world leader enterprise grade data quality into the big data environments. So last week we announced our Trillium Quality for Big Data product. This product brings unmatched capabilities of data validation, cleansing, enrichment, and matching, fuzzy matching to the data lake. We are also leveraging our Intelligent eXecution engine that we developed for data integration product, the MX8. So we are enabling the organizations to take this data quality offering, whether it's in Hadoop, MapReduce or Apache Spark, whichever computer framework it's going to be in the future. So we are very excited about that now. >> Congratulations, you mentioned the data lake being a swamp, that Dave Vellante referred to. It's interesting, because how does it become a swamp if it's a silo, right? We've seen data silos being antithesis to governance, it challenges, certainly IoT. Then you've got the complication of geopolitical borders, you mentioned that earlier. So you still got to integrate the data, you need data quality, which has been around for a while but now it's more complex. What specifically about the cleansing and the quality of the data that's more important now in the landscape now? Is it those factors, are that the drivers of the challenges today and what's the opportunity for customers, how do they figure this out? >> Complexity is because of many different factors. Some of it from being global. Every business is trying to have global presence, and the data is originating from web, from mobile, from many different data sets, and if we just take a simple address, these address formats are different in every single country. Trillium Quality for Big Data, we support over 150 postal data from different countries, and data enrichment with this data. So it becomes really complex, because you have to deal with different types of data from different countries, and the matching also becomes very difficult, whether it's John Furrier, J Furrier, John Currier, you have to be >> All my handles on Twitter, knowing that's about. (Tendu laughs) >> All of the handles you have. Every business is trying to have a better targeting in terms of offering product and understanding the single and one and only John Furrier as a customer. That creates a complexity, and any data management and data processing challenge, the variety of data and the speed that data is really being populated is higher than ever we have observed. >> Hold on Jim, I want to get Jim involved in this one conversation, 'cause I want to just make sure those guys can get settled in on, and adjust your microphone there. Jim, she's bringing up a good point, I want you to weigh in just to kind of add to the conversation and take it in the direction of where the automation's happening. If you look at what Tendu's saying as to complexity is going to have an opportunity in software. Machine learning, root-level cleanliness can be automated, because Facebook and others have shown that you can apply machine learning and techniques to the volume of data. No human can get at all the nuances. How is that impacting the data platforms and some of the tooling out there, in your opinion? >> Yeah well, much of the issue, one of the core issues is where do you place the data matching and data cleansing logic or execution in this distributed infrastructure. At the source, in the cloud, at the consumer level in terms of rolling up the disparate versions of data into a common view. So by acquiring a very strong, well-established reputable brand in data cleansing, Trillium, as Syncsort has done, a great service to your portfolio, to your customers. You know, Trillium is well known for offering lots of options in terms of where to configure the logic, where to deploy it within distributed hybrid architectures. Give us a sense for going forward the range of options you're going to be providing with for customers on where to place the cleansing and matching logic. How you're going to support, Syncsort, a flexible workflows in terms of curation of the data and so forth, because the curation cycle for data is critically important, the stewardship. So how do you plan to address all of that going forward in your product portfolio, Tendu? >> Thank you for asking the question, Jim, because that's exactly the challenge that we hear from our customers, especially from larger enterprise and financial services, banking and insurance. So our plan is our actually next upcoming release end of the year, is targeting very flexible deployment. Flexible deployment in the sense that you might be creating, when you understand the data and create the business rules and said what kind of matching and enrichment that you'll be performing on the data sets, you can actually have those business rules executed in the source of the data or in the data lake or switch between the source and the enterprise data lake that you are creating. That flexibility is what we are targeting, that's one area. On the data creation side, we see these percentages, 80% of data stewards' time is spent on data prep, data creation and data cleansing, and it is actually really a very high percentage. From our customers we see this still being a challenge. One area that we started investing is using the machine learning to understand the data, and using that discovery of the data capabilities we currently have to make recommendations what those business rules can be, or what kind of data validation and cleansing and matching might be required. So that's an area that we will be investing. >> Are you contemplating in terms of incorporating in your product portfolio, using machine learning to drive a sort of, the term I like to use is recommendation engine, that presents recommendations to the data stewards, human beings, about different data schemas or different ways of matching the data, different ways of, the optimal way of reconciling different versions of customer data. So is there going to be like a recommendation engine of that sort >> It's going to be >> In line with your >> That's what our plan currently recommendations so the users can opt to apply or not, or to modify them, because sometimes when you go too far with automation you still need some human intervention in making these decisions because you might be operating on a sample of data versus the full data set, and you may actually have to infuse some human understanding and insight as well. So our plan is to make as a recommendation in the first phase at least, that's what we are planning. And when we look at the portfolio of the products and our CEO Josh is actually today was also in theCUBE, part of Splunk .conf. We have acquisitions happening, we have organic innovation that's happening, and we really try to stay focused in terms of how do we create more value from your data, and how do we increase the business serviceability, whether it's with our Ironstream product, we made an announcement this week, Ironstream transaction tracing to create more visibility to application performance and more visibility to IT operations, for example when you make a payment with your mobile, you might be having problem and you want to be able to trace back to the back end, which is usually a legacy mainframe environment, or whether you are populating the data lake and you want to keep the data in sync and fresh with the data source, and apply the change as a CDC, or whether you are making that data from raw data set to more consumable data by creating the trusted, high quality data set. We are very much focused on creating more value and bigger insights out of the data sets. >> And Josh'll be on tomorrow, so folks watching, we're going to get the business perspective. I have some pointed questions I'm going to ask him, but I'll take one of the questions I was going to ask him but I want to get your response from a technical perspective as CTO. As Syncsort continues your journey, you keep on adding more and more things, it's been quite impressive, you guys done a great job, >> Tendu: Thank you. >> We enjoy covering the success there, watching you guys really evolve. What is the value proposition for Syncsort today, technically? If you go in, talk to a customer, and prospective new customer, why Syncsort, what's the enabling value that you're providing under the hood, technically for customers? >> We are enabling our customers to access and integrate data sets in a trusted manner. So we are ultimately liberating the data from all of the enterprise data stores, and making that data consumable in a trusted manner. And everything we provide in that data management stack, is about making data available, making data accessible and integrated the modern data architecture, bridging the gap between those legacy environments and the modern data architecture. And it becomes really a big challenge because this is a cross-platform play. It is not a single environment that enterprises are working with. Hadoop is real now, right? Hadoop is in the center of data warehouse architecture, and whether it's on-premise or in the cloud, there is also a big trend about the cloud. >> And certainly batch, they own the batch thing. >> Yeah, and as part of that, it becomes very important to be able to leverage the existing data assets in the enterprise, and that requires an understanding of the legacy data stores, and existing infrastructure, and existing data warehouse attributes. >> John: And you guys say you provide that. >> We provide that and that's our baby and provide that in enterprise grade manner. >> Hold on Jim, one second, just let her finish the thought. Okay, so given that, okay, cool you got that out there. What's the problem that you're solving for customers today? What's the big problem in the enterprise and in the data world today that you address? >> I want to have a single view of my data, and whether that data is originating on the mobile or that data is originating on the mainframe, or in the legacy data warehouse, and we provide that single view in a trusted manner. >> When you mentioned Ironstream, that reminded me that one of the core things that we're seeing in Wikibon in terms of, IT operations is increasingly being automated through AI, some call it AI ops and whatnot, we're going deeper on the research there. Ironstream, by bringing mainframe and transactional data, like the use case you brought in was IT operations data, into a data lake alongside machine data that you might source from the internet of things and so forth. Seem to me that that's a great enabler potentially for Syncsort if it wished to play your solutions or position them into IT operations as an enabler, leveraging your machine learning investments to build more automated anomaly detection and remediation into your capabilities. What are your thoughts? Is that where you're going or do you see it as an opportunity, AI for IT ops, for Syncsort going forward? >> Absolutely. We target use cases around IT operations and application performance. We integrate with Splunk ITSI, and we also provide this data available in the big data analytics platforms. So those are really application performance and IT operations are the main uses cases we target, and as part of the advanced analytics platform, for example, we can correlate that data set with other machine data that's originating in other platforms in the enterprise. Nobody's looking at what's happening on mainframe or what's happening in my Hadoop cluster or what's happening on my VMware environment, right. They want to correlate the data that's closed platform, and that's one of the biggest values we bring, whether it's on the machine data, or on the application data. >> Yeah, that's quite a differentiator for you. >> Tendu, thanks for coming on theCUBE, great to see you. Congratulations on your success. Thanks for sharing. >> Thank you. >> Okay, CUBE coverage here in BigData NYC, exclusive coverage of our event, BigData NYC, in conjunction with Strata Hadoop right around the corner. This is our annual event for SiliconANGLE, and theCUBE and Wikibon. I'm John Furrier, with Jim Kobielus, who's our analyst at Wikibon on big data. Peter Burris has been on theCUBE, he's here as well. Big three days of wall-to-wall coverage on what's happening in the data world. This is theCUBE, thanks for watching, be right back with more after this short break.