 Okay, we're back here, live at Strata. This is siliconangle.com. This is our cube, our broadcast, where we go out to the events, talk to the smartest people we can find, and we don't care who they are. They could be entrepreneurs. They could work at Bank of America. They could be anywhere. As long as they have knowledge and they can extract a signal from the noise with us, we want to talk to them. My co-host is Dave Vellante, the founder of wikibon.org, our research team, putting out amazing, the first study on big data in the history of the big data industry. Go to wikibon.org slash big data and you will find the market sizing and revenues by vendor in the big data space. This is a comprehensive report. Go there, check that out. This is our cube and our first guest is a longtime cube alumni, Abhimeda, who is the founder of Strata and also was at Bank of America at the original Hadoop world. Been with us at the cubes, present at creation in 2010 and launched his company here on the cube. Our show. Exactly a year ago. A year ago here at Strata, Abhi, welcome back. You coined the phrase, the industrial revolution, the data factories, all is happening. You're a visionary, you're predicting the future, you're an entrepreneur, great guy. Well thank you. One of our favorite guests, welcome back. This is mutual love here, I love it, this is great. But I won your anniversary guys and we launched Trasada, as you said, exactly here on Strata a year ago and boy, has Strata changed, the cube has grown. Trasada has its product out in production with a client, so, and to your point, I think the predictions we had made and what seemed at the time, bold statements are all coming true. When you have 2,000 people sitting in a room talking to visionaries, I now call them revolutionaries on what this revolution can enable. It's exciting, it's very exciting times. Yeah, well when we first met at Hadoop World in 2010, I had been in Dallas, I've gotten the plane, left storage networking world, it was the best thing I ever did. The big data's exploded, you could just feel from the vibe and particularly from the interview we did with you, how real big data was and what it was going to mean in terms of the way it was going to change so much of the, not only the IT industry but society in general. That's correct, I did a keynote today morning talking about the power of the same thing because I think we are finally realizing there's a dimension to data and data analysis that has a power and a transformative power to solve problems we literally couldn't solve before and it's something that people have said multiple times but I think we are finally seeing that notion transcend from just the web to other industry verticals and tackle issues like healthcare or looking at the real estate mess globally which is a $14 trillion problem and offer creative ideas but also tangible solutions at a scale that just literally wasn't possible before so it's a very exciting time. So Abhi, let me ask you a question, I want to drill down into an area because you've been a participant in the industry at many levels but also pioneering with your startup so you're kind of wearing multiple hats, kind of pioneer, entrepreneur, business owner now and so on and so forth but we were just talking earlier today and yesterday about the stages of the big data business in 2010 it was, hey what does it do when you had guys like Alpha Geeks like us saying oh man this is a vision, it's going to happen, it's a future data factory so there was a conversation of one camp saying man some great stuff happening, we see the future and then the other camp was like what is it do again and so 2011 became, hey this is big data industry, it's actually a bona fide business model. Right. Right so companies are being formed. Absolutely. Investments taking place. 2012 we're seeing platform maturity and growth in applications. Absolutely. Dave and I were predicting 2013 being when it starts raining money meaning value to customers. Agree. Do you agree with that? Absolutely, I use these terms for in a very similar terminology and I'll answer the question in three short parts. First is let's look at the errors, right? So 2011 and I'll break the errors in 12 months because errors are no longer decades. I'm calling 2011 the year of validation and education so every single big tech company we know came out and acknowledged Hadoop, right? They all said. I'm in. I'm in, exactly, I get it. This is no longer a web enabled or web driven platform only for use for web data. They all came out and said no, this is gonna be a real transformational enterprise technology, which I think was big. I'm seeing 2012 as the year of scaled up testing. So this is real testing, scaled up proof of concepts in every single industry vertical you can think of. That's 2012. And 2013. Scaled up proof of concept, just to kind of clarify that. Most people think a proof of concept has a little bit of test money. You're talking a little bit different. Absolutely, good point. Bigger scale, talk about the scale of proof of concept. It's a very good question. I think that there are two dimensions to it. First of all, the proof of concept is no longer around let's build 10 nodes and test if I can install Hadoop, right? I think we are beyond that phase. The proof of concepts are more about I know I have problems to solve that I cannot solve on existing technology stacks or technology tools. So can you help me apply in a proof of concept this new technology platform and show me what problems can I solve? So I can prove out the return on investment to make the big leap and completely re-architect the enterprise data stack around Hadoop. So I'm calling 2013. And the leap is the complete reconstruction of their business model. Complete reconstruction of the business model. I think a realization, but also the complete reconstruction of the data infrastructure within an enterprise. I made a statement today morning, which is the data stack, as we knew it, has been completely commoditized. The last remaining piece was data intelligence, right? Or what we used to call BI. And that's now commoditized. So that means technology is no longer the constraint. Technology is no longer the constraint. Exactly, the point I made, technology is no longer the constraint. So if technology is no longer the constraint and you have petabytes of data and you have problems that need intelligence at a very granular function, to do bottoms-up analysis, not top-down, to solve problems that could range from understanding the spread of disease to understanding how you price your credit cards, it can now be done. So I think 2013, we will see to reinstate your point, re-emphasize your point. 2013 will be the year where we will see entire data infrastructures being built around Hadoop in industry verticals that we hitherto would have thought would not have done it. So let me ask you a question. So a lot of money pouring in. Yeah, I think, I do agree. I think so too. Raining money means you start to see the real value of dollars. We were talking earlier about the role of virtualization plays and flash is playing in the construction of database, the gear that powers the big data. Dave and I were talking about VMware's lack of big data strategy. Do you have any perspective on that? Because they have yet to come out with their big data strategy. And essentially, there's a lot of things that they could do with it, with data layers. So the data layer is commoditized, but where does it fit? So close to the server, we're seeing with flash, certainly you can put it close to the server. So virtualization VMware owns the enterprise. Flash is now making a massive inroad with Fusion IO and others in the enterprise. Where's the big data strategy from VMware? Do you have any perspective on that? Yeah, so I think I'll make it a lot slightly broader than VMware specifically. I'm not... The virtualization players. Yeah, so virtualization players in general. So I think the biggest, I don't think the best get the biggest secret of the world of big data that people don't talk enough about and we are very passionate about is the fact that there really is not a analytics cloud in the market. So when we go to our customers, if you want to be able to do analytics, at scale, big deep data processing and then deep analytics in the cloud, you cannot do that. Because think of it, right? You look at an Amazon and this is deep respect for what Amazon's done. But I can't put a petabyte of data on S3 and then move it to elastic map and produce to process it. And then you've got IO problems too. You've got IO problems. Which goes to the second point, you cannot charge customers by IO in the world of big data analytics. Because that's like giving somebody a beautiful, yeah, putting a great lunch buffet in front of you and trying to hand it in the bag and say, you know what, only if you pay me for that food item and that food item separately, I'll let you taste it. So there is a massive gap. Virtualization as a larger industry, there's a massive gap in the market of being able to do what we are not calling big data as a service. The infrastructure part of that big data as a service is a massive whole. So I tell every single VC I meet, if you find a company that can enable you to do analytics in a virtualized fashion, not in a virtualized architecture, virtualized fashion, but allowing you to leverage it with data as local, invest in them because that is a massive whole in the market. So that's one. I think on the second part though, because I've always said this, big tech, the only way big tech innovates is by buying companies, right? Innovation means how big of a checkbook do you have, right? So I think there is, I'm calling it, the next, we're seeing a Darwinian revolution for big tech. And it's gonna last for the next 50 years, but I think the next five years, we will see a incredible amount of M&A frenzy in this space where finally the action, and I think three of us were early in this in calling it, the action is squarely gonna be in the application tier, not in the infrastructure tier, because it's commoditized, and we will see an incredible frenzy of M&A amongst the big tech guys to get there. So I think, can VMware figure out a path to leveraging the big data trend to re-energize the business? I think it's going to rest on how smart of an M&A guy or team they have to figure it out for them. It's such a great use of cash, you know. Paul Moritz is definitely smart enough, sure, but the question is speed, and so they bought spring source, obviously, at that point, arguably a big number, looking good right now with the framework growth. They have a huge hole, and they had to fill that, so I don't think VMware can figure it out. Personally, I'm not sure, so, I mean, that's your opinion as well. Maybe they'll buy Cloudera, because you know, a lot of the guys from VMware left VMware to go to Cloudera. I actually did not know that, that's an interesting insight, you know? You like Collins at VMware? I like you, John, because you make bold predictions, and life is short, you got to make bold predictions. I'm my own boss, no one can fire me, so. Which is the beauty of being an entrepreneur, right? Mandel, Rosenblum, and also Diane Greener, Investors in Cloudera, you have a lot of DNA from the virtualization space, Omar Awadallah, did a PhD in virtualization under Mandel, so there's a lot of similarities and culture there around big data and virtualization. I think it's an interesting point, and I'm going to make, this is as far as I'll venture in my thought, which is the industry today, in the world of big data, being the fact that we play in the application tier, and we certify our platform in every single flavor of Apache, which I can tell you is not very different, and the performance, nothing's different, right? It's a different branding, and I think it's unclear, and it's important for the market, there needs to be a clear winner in the short, if Hadoop is to be and hold the promise it has, which is to be the data platform for the future, it is in all of our best interests to coalesce around one player who can be the red hat of the, of the, Get certainty, get certainty into the, get certainty into the, take that risk out of the equation. Because I think it's becoming very, the Hadoop, and you've heard me say that, Hadoop was always enterprise ready, but as we are now transitioning into use cases and business use cases with our clients, the clients are going to start driving what the open source community needs to build, and I can tell you, I can tell you with 100% conference today, what's on the roadmap is not what the clients are asking for, and I think that needs to very quickly coalesce around making sure that making Hadoop scale to be the enterprise data platform needs things to be fixed, that the enterprise needs to inform. But it's not happening right now, it's Darwinian, right? It's not happening. So Abhi, because we make bold predictions and you also make bold comments that turn out to be true. We both are geniuses, I guess, self-promotion each other. If you're a VC, and you're, say, Frank Artali at Ignition, or Ping Li at Excel, two of the smartest VCs I know in the business in big data, you know, Mike Olson talks about application surge. You guys are doing some good work there. If you're a VC, if we're a VC partnership, three of us, what investments would we make on the application side? What would we look for? So, obviously, as applications come out, which ones are fundable? Because there's a lot of apps that are lifestyle businesses. You have SaaS is really easy to deploy. You could have essentially ISVs, data ISVs, all kinds of new nuances of business models. But we're VCs, we want to invest in big wins. What are the big apps? What would we have to look for? So, I think it's, I'll take Prasada's example, because I think it's close to my heart, but it's a model that's now worked in the market because we're now in production live with clients. I think there are three key requirements for an app, a big data app, to be successful. Number one, you have to understand that big data has two core pieces. There's the data processing end and the deep analytics end. And if an application only does one, and not the other, I would not invest in it. Because the reality is you can't do deep analytics unless the data has been processed, right? So, you've got to be able to find a analytical platform that can do both those things. That's number one. Number two, you've got to be able to identify and build your application platform to be API ready. So, because here's the reality of it. No matter how many, how much resources collectively VSVCs may invest or ISVs data may invest, the application ecosystem will always outrun and outpace the application platform. So, you've got to make sure that what, as you're writing the platform, is API ready to enable the SIs, to enable the other players to write their own applications on top of it. That's point number two. And point number three, and again, we've been bold on this prediction earlier, but I think we're seeing it happen as we speak. We will absolutely see the verticalization of applications analytics around data types. So, for financial data, there is only one plan in the market. It's called Truseta, which you guys love, right? And you'll see similar coalescing of analytical platforms around other data types. And I've changed that slightly from industry verticals. So, I think the way we define industry verticals has to change. We have to think more of data verticals and industry verticals. Because guess what? Retailers, and we've been getting tons of inbound calls given our application from retailers. And we used to scratch our head and go, why? I'm a big data for banking company, right? And the reality is 80% of all data in a retailer is financial data. The problem they're trying to solve is what my application solves, total view of customer. So, we have changed our, but I think those are three things that we would watch out for on the big data analytics side from an application perspective. Great. So, thou shalt, thou shalt not just process, thou shalt do deep analytics, thou shalt do the API. Or vice versa, thou shalt not only do deep analytics, but thou shalt also do the process. And then thou shalt go vertical. Okay, talk about, let's talk about something, you're talking about, I asked a VMware question, you kind of made it a little broader about virtualization. So, let me ask you about Hadoop versus other platforms like Cassandra and this new open source project, we're seeing it even on the IO side, Dave and I are talking about Node.js being fundamentally on the IO side. So, you got a lot of different pieces to the puzzle. Correct. Okay, so talk about how you're seeing those evolve. So, I think I have actually a very simple answer to that one. I call it the HDFS ecosystem, not the Hadoop ecosystem. And if you're not on the HDFS ecosystem, good luck. So, we are a company completely built on the HDFS ecosystem. I think the HDFS ecosystem will evolve very rapidly to address pretty much every single analytical process you would like to run, including and including things like real time and streaming. We actually already enable real time analytics on our platform. Streaming is an app that will absolutely get built as a horizontal application on top of HDFS ecosystem. So, I think as far as we're concerned, the way we look at the market, the HDFS ecosystem is the winner. The HDFS ecosystem will be the data path for the future. And while, let's take the example you used, while I love Cassandra for some specific things it can do, the thing I do not like about Cassandra and what they did was write their own file system. So, if Cassandra had actually leveraged the HDFS file system, it would have been great. Here's why I say that. We're seeing these interesting tools develop, but they're breaking the golden principle of big data. They're asking you to duplicate the data, which is a massive problem. So, let's take a great example. Lucene, we are big believers in the next phase. If you were to ask me the question differently, what would be the next two or three projects we believe are gonna be massive in the world of big data? We are big believers that text-based search on top of deep analytics is gonna be the killer app, which immediately takes your mind to Lucene. But here's what I don't like about Lucene. In Lucene today, in Lucene Solar, I gotta duplicate the data. I'm not gonna do that. I'm just literally, I refuse to do that. So, I think, same thing goes for R. We are massive believers in R, but in R today, R does not work on HDFS. R works on a very small implementation of HDFS, and Revolution will tell you that. So, we're working with Revolution to actually make R work at scale in the HDFS ecosystem. So, I think, if the newer things are not being built in the HDFS ecosystem, good luck. So, ripping out HDFS and dropping in Cassandra doesn't do it for you. That's just, you feel like the wrong approach. Exactly right. Cassandra, if it worked native HDFS, it would have been a lot more powerful. I think we'd have seen a lot more adoption in the market as well. And there's a lot of talk about machine learning at this event. Talk about that a little bit. So, I think machine learning has two dimensions. I'll break it back to the dimensions that we mentioned. There's a data processing side of machine learning, and then the deep analytics side of machine learning. I think we all get enamored with when we hear the word machine learning, and we jump into the deep analytics, right? Because it's sexier. Sure, right. But I think, and if you're a great example of it, Tracita has written as many proprietary algorithms in machine learning on data processing, probably more as we have on deep analytics. Because here's the problem, and here's the power. The ability to actually take massive amounts of data, both structured and unstructured, and then have it available for analytics, which is a massive processing job, right? You got to ingest it, clean it, de-dupe it, merge it, parse it, then rerun it, then run the analytics on it, is a machine learning problem. And the way you address those issues, especially when you bring in structured and structured together, which is another big cow I have. I don't agree with this concept of do structured and relational, and do unstructured, and do. That's bullshit. Unify them. You've got to unify them, too, because structured by itself, interesting. We've been doing it for 50 years, and we have to have massive problems to solve. Unstructured by itself, yeah, kind of interesting. Bring the two together. Killer. Or killer. Transformative, right? Okay, so I want to ask you about H-Base. Obviously, we'll be at H-Base conference. We run H-Base. Obviously, H-DFS, you talked about the killer file system. So what's your view on H-Base? How do you feel about it? The growth of H-Base, their challenges, and opportunities, what do you think about it? So I think H-Base is a part of the H-DFS, a platform we love. So I think H-Base becomes over time, and I think the techies will hate me for me saying it, but I'll say it anyway, because it's a business perspective. I think H-Base becomes the database on H-DFS, long-term. From a perspective of what? What's wrong with that? I don't know, techies don't like that, for some reason, techies don't like that. But I think so, big believers in H-Base. Yeah, because you're forcing it on. Sounds good to me. Because you're forcing it on. I'm forcing it on them, this is accurate. But I think long-term, H-Base becomes a massive disruptor in the kind of functionality users are expecting from a database you can expect from H-Base. And it's going to be incredibly powerful. Especially for solutions or use cases where you want very, very quick, random write and read, which there is no better solution for it. So H-Base is going to be a massive winner. We use H-Base a lot internally. You're using it? We use H-Base a lot internally. So we use it, perfect. So I think H-Base is definitely one of the winners. It's got to do some work. What about the challenges? Where do they need to improve H-Base? I mean, honestly, it's not really mainstream enough. In my mind, mainstream developers, you've still got to be pretty strong. CS due to H-Base, what you see is the opportunities, challenges. I think I have a different perspective on it. I think because the H-Base ecosystem is an open-source ecosystem, we should not expect the open-source community to solve all the problems. Because they won't, right? Because a smart geek or a smart quant will do what he wants to do. And we should let them do that because that is the very spur of big innovation. But I think you will see companies like ours or Vividata innovate around the so-called imperfect pieces of open-source technology and make it more user-friendly. Or build the bells and whistles you need to make it effective in the vertical you choose. So I think, slightly biased because I'm in the business of setting my own platform. But I think you would see H-Base works today. It's got its core, but it works today. And I think you would see companies around the open-source ecosystem making it more user-friendly. Okay, so I have to ask you about some things that you've mentioned two years ago in theCUBE about data factories and how disruptive that will be. We just had on theCUBE earlier the Metropolitan Chicago Information Center, Virginia Carlson, the president, who's solving a lot of problems as non-profit. They use data to solve, try to use big data to solve problems around health care deployment for federally funded people. So in your keynote today you talked about big data and socioeconomic issues. The question that we are exploring on that kind of impact of society question is people hoarding their data. So we also heard Tim Estes from Digital Reasoning who we love talking about he doesn't trust Facebook. So people commercially are trying to grab all the consumer data to sell it back to the companies for fees. Tim was worse actually, he said he trusts the government more than he trusts Facebook. To me that's Tim Estes. Oh my God, that says it all. Yeah, that's how fucked up Facebook is right now. And I said, how about Google pause? He says, are you kidding me? I guess that's ranked slower than Facebook. And by the way, Facebook's all ex-Google. So again, you see where the DNA is. Anyway, I get distracted. You guys have data, but you're a privately held company. So where are we with this notion of opening up the data so that combining private data and government data together, if we're going to have data factories and have an industrial revolution like impact, where do we need to be with data? We talked about the tech and getting the weeds in the tech. Let's talk about the society level. Where do we need to be mindset wise and policy wise with data? So I think, first of all, great question and probably a trillion dollar question because I don't think there's a good answer to it but I'll give you my perspective on it. I think we need to, as an industry, and I mentioned this to Jeff Kelly who's a phenomenal contributor on Wikibon and someone I respect a lot. I mentioned this to Jeff as well. We need to, as an industry, stop belittling data. Data is an important asset. It also is sometimes a very private asset and we need to understand and recognize that. The conversation we are not having that Truseta is taking the lead, just given the industry vertical we are on, is around data privacy and that it has to get addressed, John, to answer your question in a complete way. So I think there needs to be a conversation around data privacy, data identity, and data ownership. Before it becomes clear what data can be used for what problem and what format and what way and who benefits from it. Okay, so let's go. That's the whole key. Who benefits from it, right, sir? Good escape clause answering the question because I have you on record two years ago at the data factories, but let's me bring to what we think is the hottest trend right now that's in the industry is identity and trust. So apply identity and trust to data. So that's going to be a big component in the privacy question, so what's your perspective on identity and trust? So no escape clause, I was going to answer your question. I was just busting your tongue. So I think data is, there is no choice for us. There are two massive trends that are coalescing, right? And I hate to use both those terms because they've now suddenly gone from sexy to just being overused, but cloud and big data. I think cloud is no longer a question on if the cloud happens. The cloud will absolutely happens. The question has become when does the cloud happen, right? Which in turn drives the conversation around data factories. So I'm going to data factories because if the cloud happens, right, sorry, when the cloud happens, the data is going to sit outside your own, what do you thought where your own four walls? Is that necessarily bad? The answer is no. Probably data is most secure with a company that is in the business of managing data, processing data, probably, right? It depends on it, yeah. Processing data, then in a company that's core competencies, banking or health care. Oh, yeah, oh, in the cloud. Oh, yeah, yeah, I thought you were saying in reverse. Oh, yeah, I totally agree. Because that's a core competence, right? That's the first thing. It's a differentiator for those guys. Absolutely. I invested it. So, and the thing that the cloud has been missing and being from an industry, everybody hates the word cloud and they shut the doors and windows when the clouds appear, right? But leaving that aside, leaving that aside, I think big data becomes a trend that makes the cloud enterprise ready. And when that happens, to answer your question specifically, I think we'll have no choice but to live in an ubiquitous data world where data, by its very nature and definition, is open and allows for innovation to solve some white hat problem. So what is missing? What is missing is the cloud infrastructure. What is missing is, and some of it is just timing, what is missing is the applications, the white hat applications. No one has come to me, and I've mentioned this to Jeff, again, thing. We need to think big, we need to dream big about big data. I love coming over here because you guys force me to dream big about big data. Where are the conversations that people are sitting and saying, if I combine Facebook data with Google Web data, with GPS information, I can eliminate fraud in the financial system. Who is having that conversation? You are. Nobody. You are. You are. So you guys are solving that problem. That's correct. So I think we need to identify two or three white hat, big ticket items, put them on the board and say, if we had data ubiquity, open data access, I can solve massive trillion dollar problems. And I think you will see when that happens, the government, because the government will be, whether we like it or not, the government will be involved, as this should be. You will see private enterprise, and you will see startups coalesce around this notion of data ubiquity, data openness, and go after it. And that's the privacy issue. The problem is not just a public policy issue. It's a platform, suppliers. It's all those stakeholders. Because if the government came to three of us and said, okay guys, I want it to be open. Can you manage security privacy? Can you manage user provisioning and access? The answer is, well, not really. I would love to, not yet, right? And then I think there's, someone you need to bring to theCUBE that I promise I'll bring next time is a guy called Sandy Pentland. He's a partner and chief privacy officer at Triseta. And he laid out a thing called the new deal on data to answer your question. Is it published? Is it published? It's published. It's at Davos. Let's get it republished on SiliconANGLE. Perfect. And we'll do that and get him on theCUBE. So I love chatting with you because we can be philosophical. We can also talk about tech and cool stuff. You're one smart CUBE alum and we're psyched to have you on. But I want to ask you now a philosophical question that's a little bit more kind of getting the clouds. So big data is a disruptive marketplace, obviously. And it's like we're all climbing the mountain and when we get to the top of the mountain, whoever gets there first can look at the vista and see the valley of opportunities, wealth, creation, benefits to society. At the top of that mountain, you look down on that valley of opportunity. What do you see and what do you go after first? If you're an entrepreneur or yourself? I mean, what's your perspective? I mean, new opportunities are emerging around big data that rival the PC industry back in the late 70s, early 80s. So this is going to be a massive shift. What's that flourishing land look like for you? The flourishing land is massive. I mean, you guys wrote a great report, by the way, which I enjoyed reading. And I think putting a number out there is important for the industry. Big data is not a $50 billion market. It's probably a $50 trillion market. Yeah, it's much bigger. I agree. Because it's going to reshape and resize to answer your question. You're not even going to be able to count it, right? I mean, you're not even going to be able to count it. Because the amount of innovation, a platform like Preseta can enable for a bank to literally rebuild the banking system on, what is the number of that? $50 trillion, right? The mortgage market is $14 trillion. So I think I have a very simple... Gabe, go back to the drawing board and your market sizing. But I have a very simple rule to look at the opportunity or visualize the meta. And it is, the successful companies of the future are not going to worry about the tools. Are not going to worry about the widgets. Are not going to worry about real time versus batch. They're only going to worry about one problem. What business problems can be solved that we could not solve before? And whoever is coming to you and has told you that they've found the business problem, whether it's in healthcare, in retail, in telecom, in the web, in social or in financial services, that meta skies the limit. Because I think not only will you see people like us making money in M&A, I think you will see the birth of the next generation technology data companies. But the line between technology, data and business is completely blurred. It's one vast ocean and the opportunity is truly limitless. So that's that one simple rule. What business problem are you going to solve? Okay, so I need to kind of go the next step, which is for sure I think it's massive. I'm excited by it. And I'm just, I'm not just saying that to pump up the marketplace. I think we are seeing predictive and real time analytics and data being a big part that no one ever had that conversation in data warehousing and business intelligence in the past. And the lines are so blurred. The business, the technology, what is and what is. I mean, it's all big data. That's right. I guess my question then is about the new user experience. So assuming it's a $50 trillion market or more, hundreds of trillions of dollars, big data will enable. What is the user experience going to be like? We had, again, we had the folks from Digital Reasoning on talking about we use our mobile phone because the data management's not that good. So you got to be connected to the data all the time and that new filters are going to come along. So all these new stuff you can dream up the future. What is the expectation of these users going to be? What is your vision and perspective on the user experience? Great point. The environment, the UI, the real life experience, what's your perspective? So here's the second actually dark secret in big data. Big data does not take the human element away from decision making, which is going to answer your question. I believe, and so on our platform, to use Trissida's big data platform, you don't need to learn MapReduce. You actually don't even need to know how to use Excel sheets. You press buttons. So we believe that the way we break out the market, data processing with deep analytics will be fully automated, right? Those pipelines will absolutely be fully automated because you finally can using artificial intelligence, machine learning, and the new expressive power of MapReduce and HDFS automate those two pipelines. The user experience in that case, John, very simply becomes a business user. And here's how I define the business user. A user who is in touch with the direct and customer. So it could be anybody. It could be a teller at a branch. It could be a checkout lady at the store, or it could be the person sending a cell phone at the AT&T store, right? That person has ubiquitous real time access, not to data, but to insight. In a user friendly way, but all you have to do, like on the iPhone, is swipe something, touch something, see what the information is, and make a decision. So I think the user experience becomes extremely, extremely centric, looking more like the consumer apps, rather than the enterprise apps, where you need the act of God and 5,000 engineers to write a report, and it takes six months. I think that completely disappears. I think you'll see a massive consumerization of the user experience. If you actually want to show you our platform, our platform has something called tiles. The tiles are codified by business problem. And literally, all you have to do to make it use is press a button. And when you don't like, or you have to change the cycle, you move some Lego blocks around, and voila, you can start another business problem. So I think you will see people, like you and me, spending a lot more time accessing the information, and making better decisions for their customers in a consumer-like way, from the user experience, rather than worrying about writing SQL codes. So pushing that decision, making way out and distributing affection, and then still, you need the human to do the last, maybe it's 100 feet, maybe it's not the last mile. Is that correct? Okay, Abhi, you're a CUBE alumni. You've been a regular contributor now, I guess, on theCUBE. Great perspective. Now CEO and founder of your own company. Visionary, thank you so much for your perspective. As usual, great segment. Numbers are up. We can know when content's good, people start to do it. The counter goes up, so. I think you're also looking, the biggest thing in order is the change in strata. The first time I spoke here last year, there was no makeup. This time they put me down and powered me down before I went for my keynote, so. I think that's a sign of success. You're one handsome man, I gotta say. Thank you, ladies out there. No makeup on theCUBE, not yet anyway. Congratulations. John, always a pet, thank you so much.