 Okay, we're back live here in SiliconANGLE.tv, theCube, our flagship telecast, we go out to the event, extract a sneak from the news, I'm here with the Dean of Big Data, who we, the CUBE alumni, Bill Schmarzo, a good friend with EMC, Palo Alto, native with me in there. The Dean of Big Data, we coined that term just kind of off the cuff, but it sticks for you because you know, you're very professorial, you're talking to customers, you're on the whiteboard, you got your own room at EMC, so tell us what's going on with you right now, what's going on in the marketplace. So, I'm still spending almost all my time talking to customers, working with them, trying to figure out where and how to start. The conversations with them have changed dramatically in the last 12 months. We've gone from having conversations where we're trying to explain that CUBE's not necessarily Big Data on itself, right? So we're having conversations more about business enablement, about business transformation. I had an interesting conversation with a large grocery chain recently about how they can leverage all their wealth of customer loyalty data and their social media data to dramatically improve the shopping experience of their shoppers, right? So the level of conversations are becoming a lot more real and a lot less theoretical. So in terms of what's happened one year from now, one year ago to now, what in your mind describe your feeling of the market? And we're at Oracle Open World. So go back one year to Oracle Open World last year, you were on the CUBE talking to with us. What's happened this year? And what's the mood within the Oracle ecosystem? Well, I think that the mood has gone from one of being just education, just people trying to figure out what it is. To now, people are starting to take very serious look at how I can leverage both the wealth of data I have access to in my organization, both inside my organization and outside the four walls of the organization, and how then all these new technologies like Hadoop and NoSQL and MPP architectures, how those technologies help me to tease more insights out of those data. So there's, I think within the Oracle community and across, I mean, all the communities of people we're talking to, much more realization that there's value there and it's not necessarily a technology discussion that technology is a part of that discussion, but technology is the enabler. It's not the discussion, it's only the enabler. So we just put out some new research as a sample. We have a lot of other kind of cross sections of markets and verticals, but if you go to slideshare.com slash SiliconANGLE, you'll see our first big data report called Trend Connect where we've sampled about 180,000 people on Twitter and are classifying their interests in certain things. We're seeing that the interest for Hadoop is dropping, but the actual interest for analytics is rising, which comes back to kind of a historical perspective. If you think about it, back in the old client server days, client server, what the hell does that mean? That's like the tech. And then everything shifts to ERP, CRM, so you're seeing now an accelerated shift. It took years for that to happen in client server. Now essentially months, where now the customers are moving to solutions like analytics, do you agree with that? Oh my God, yes, I mean, I have yet to meet a CMO or a head of store operations who woke up in the morning and says, oh my God, I need Hadoop. They don't care about Hadoop. What they do care about though is the fact that Hadoop is part of the technology stack that helps me to tease insights out of all these new data sources. In fact, to me, John, to be honest with you, it's the new data sources that are the most attractive. It's like super-powered gas for your car, right? You've taken your car and now I can shove all this new powerful gas into it and whether it's social media data or sensor-generated data or data coming off my mobile systems, my mobile applications, some of our users are almost drunk with the aspect of this data that's available to them and they're realizing that the data is the gold, Hadoop and other technology are just enablers of that. Drunk, intoxicated, get the breathalyzer out, the big data. Jeremy could use that in the face of big data, his human face of big data. You're intoxicated on big data, pull them over. But people aren't intoxicated because what they're seeing is, we talk about this a lot of times on SiliconANG on theCUBE is, especially in today's market, as Joe Tucci said, CEO of EMC this morning, it's a great time, right? So in all theaters, business model, technology, marketplace exploding in innovation, there's a clear line in the street, old way, new way. And you can see who's on which side of the street. So old ways are changing fast, some need to be in the new way faster. So let's talk about that. In your mind, what are some of the new ways that people have to do it? And one of them is analytics and instrumentation and getting data. And one, knowing what to do. Well, you made a really good point there, John. You say, data and instrumentation. Now, we come from the Silicon Valley where web-based companies and web companies all are all about instrumentation, right? They instrument their websites, instrument through cookies and such. Most organizations don't think about instrumentation. They don't think naturally about how do I capture more data about my customers, my products, my campaigns? And so instrumentation to them, that's kind of a foreign word to them. So but when they start thinking about data as this economic asset, instrumentation becomes this very powerful enabler and concept, the way to get at more data more frequently. No, it sort of enables this whole high velocity thing. If you're not instrumented to capture data at the point of customer engagement, you can't have high velocity. So the whole idea of instrumentation and data as going hand in hand and part of the strategy of bringing in the raw material, the oil, right? That I think I saw some analogy that information is a new oil. It's kind of true, right? And now I'm in this mode of really trying to just get my hands and all the data I can. Just kind of, I want to bring it all into my environment even if I don't know how I'm going to use that data yet. Now we were talking in the Cube at VMworld and all the time David Blonte and I talk about this is like an observation we came away with at the last VMworld is, big data is about answering old questions faster and having the ability to ask new questions. And so with that, you know, you guys are doing this big marketing campaign the face of big data. There's not just technical questions. It's not just about the database. It's about the solutions. What areas are you seeing in your experience in the field as you architect new solutions with your customers that have the most explosive value right now? So there's actually a number of different areas, John. It's almost, it's hard to pick out any one particular area. I will say that in the healthcare space, we're seeing a lot of traction. Now healthcare's got a challenge because they've got a very diverse set of data sources and they also have some very complex privacy issues to deal with. But healthcare is one of the areas we're seeing a lot of interest and a lot of people willing to push the edge of the envelope trying to figure out how do I get access to the data in order to provide more effective care at a more effective cost, right? Better care at a cheaper cost. The other area that's starting to pop up and it's just kind of early yet is companies using data to provide a better experience. Now the experience might be, as I mentioned, the shopping experience. It could be data being used by automotive companies to provide a better driver experience. We already know that a lot of web companies have historically focused in on how do I use data and instrumentation to provide a better web experience. So the whole idea of using this data to not just tease out insights but to drive a more strategic, more engaging relationship with their customers, their partners, their consumers is also something we're starting to see a lot more interest in. So for the folks out there watching, we have Larry Ellison giving a keynote right now. We're going to cut in and out of that. Dave Vellante is actually with our team in the war room working through the analysis of this. We will have Dave on immediately when this is over. I'm here with Bill Schmarzo, CTO of EMC's consulting architectural team at EMC powering the next solutions. Whiteboarding with clients in your special den, the professor, the dean of Big Data, as we call you, a lot of experience. And you've only been at EMC how many years now? Two now, I've been here two years. So you've been in the BI space for a while, you know BI and you know data warehousing. That business is being disrupted. So I want you to walk me through the players. Give me the top five players of BI right now. Wow, that's like a totally loaded question, John. Such a typical John type of question to ask. So if you look at the traditional BI players, overall I think they're in trouble. If I take a look at, you know I used to work at business object, I was the vice president of analytic application of business objects for three years and I think as a technology they're in trouble because they're at risk of having a market move around them. The business objects, the cognosis, the hyperions, you know, micro strategy not as much so but a lot of the traditional big BI players got bought by large companies and their focus became on retaining their customer base, not on improving the technology. And as you said earlier John, people want analytics. Well I got to be honest with you, there's not much intelligence in business intelligence tools, right? They're used for reporting in dashboards, they aren't used for advanced analytics. And so the BI vendors out there who are wed to a model that focuses on reporting and dashboards and business monitoring are going to get overrun by technologies and tools that are focused in on analyzing and optimization of the business. So I think the traditional BI vendors overall are at risk of missing a very big market movement and having other players kind of sweep around them to claim that space. So let me ask you that, this is a good observation we talked about before, so what two things? What are the things that these guys need to do to stay ahead of the curve to tweak? Because the notion of having these dashboards on the front end and having structured data is not a bad thing. I mean you guys sell a variety of solutions with Hadoop and Proprietary and you're seeing Cassandra work with Hadoop and you've got MongoDB, so you have a plethora of back end new stuff to power some faster real time. But you can still have structured data in dashboards. Is it a mindset issue? Is it a technology issue or both? What do they need to do to stay ahead of the curve? And then second question is who's the new guys? And what do they look like? You don't have to name the company names, but describe the makeup of what that team might look like. So the first question, it's partially a mindset. It's a mindset of an organization that has been successful. The sales teams go to club, they make their numbers, selling reporting tools. Little motivation to change. It takes a huge amount of management fortitude to get organizations like that to make the next step. And so you have inertia working against you in those organizations. But they've also got a base technology problem, which is they don't have any advanced analytic capabilities. They don't have the ability to easily identify anomalies in the data, to identify and quantify correlations, to codify collinearity, to build linear regression and association models. They don't have that base technology. And there's a market need out there to take today's existing dashboards and move them from sort of retrospective, static dashboards to something that's more predictive and more real-time. So I think from a technology perspective and even from an architecture perspective, those companies are at risk. Now there are companies coming up, companies who are taking a look at this, who are taking a look at Hadoop. And not that Hadoop is an analytics capability. It's looking at Hadoop as a way to get at more data that can provide more fuel to help tease out insights that are buried inside. And these tools are going to start having the analytics built right into them. Think about, you know, SaaS has always been one of the dominant players in this space. And I think they're actually very well positioned to take what they have and make it easier. Make it something that your average marketing person, your brand manager could use and could easily build out an analytic model without having to know all the nuances of advanced analytics. So what's going on with EMC right now? Obviously you guys are humanizing the big data space. And how are you guys reacting? Because your marketing has been great. It's been a good air cover for a lot of the tactical stuff that you've been doing. Green Plum, what's the update on Green Plum? Well, I shouldn't speak for Green Plum. And since I'm not part of the Green Plum organization, I really can't talk about what they've done. But as a practitioner, I know what I can do with their technologies. And the focus of Green Plum to bring together, you know, structured data through their Green Plum database and unstructured data through Hadoop in the same environment off the same backplane, it gives me as a practitioner a very powerful tool I can use to really help organizations not only mind better their customer data they're already collecting, but pull in all this social media data, all this email data, all this unstructured data they have and bring it together to have a much richer view of their customers. So again, I can't speak for Green Plum, but as a solution enabler, it's a marvelous platform, I love it. So what's exciting you, if again take your EMC hat off for a second, put your big data, dean of big data hat on, what's exciting you right now about the space, that the tech business, Silicon Valley, beyond, just some of the tech involved and some of the things that you're seeing solution-wise? You know, John, I'm kind of a boring guy, right? And the thing that gets me excited is when I start having conversations with people about problems they're trying to solve, business problems. Had a conversation today in the booth about an insurance company who's trying to get down to the household rider level to predict their book of business risk, right? Something they can't do today. But something with big data they certainly can do, right? And they can model better, you know, the effect of a hurricane might have at a much granular level. They can price things differently. They can optimize their customer support and customer service. To me what's exciting is when the conversation starts moving away from technology as a science experiment to technology as an enabler for companies to solve really meaningful problems about the most important nouns of their organization, their customers, their stores, their products, their partners, their suppliers. To me, that's exciting. And I think that, you know, and I take a look across the horizon here, that's where the rubber meets the road. And I think that's what gets big data to jump across the chasm from being a science experiment that's held captive by the IT organization to be an enabler that's being embraced by the line of business as a way to gain competitive advantage. We know you've got to run at four o'clock as we're coming up on the top of the hour. Larry's finishing up his keynote. We'll have some analysis for that. But my final question for you is, what should folks expect this next year in big data? And we'll certainly see you on the queue, but the Oracle world is interested in not seeing big data now and with Cloud. I mean, there's big data mentioned but wasn't as hyped up as I had predicted. Still, it's underlying in the theme here. What should people expect of the next year between now and next Oracle open world? John, I think you made a good observation. I think that the hype has not crested yet. I think you can, you don't want to pick on Oracle, but I think Oracle is a good monitor of hype cresting. And they've not been beating the drum heavily. So I do think there's more information, more misinformation, more debate, more half truths and hearsay that's going to kind of confuse the marketplace. But I think over the next six months, nine months, we're going to start seeing real case studies of middle America companies who are embracing big data to solve some really hard, important business problems. And it's not going to be one-off things. And it's not going to be the Silicon Valley companies that invented this technology. It's going to be companies in Columbus, Ohio and Des Moines, Iowa and Kansas City, Kansas and Missouri and Albany, New York, places like that where people are starting to bring this data in and this technology to solve real problems. That's not that far away. A year from now, John, we're going to be talking about how these companies have sort of jumped over this chasm and how they've embraced the data, the technology and the, we didn't even talk about the data scientists kind of role as a way to be a real key business enabler. Do you want to, do you have a quick second to stay on? Talk about the days. Okay, let's, so then we'll stay on Larry's publishing up Dave's talk. I like the fact that I'm getting a chance to preempt at Larry here. That's good. Yeah, I mean, we've got Dave Vellante and the team in the war room right now going through all the data Larry was on earlier at CNBC defending his comments from a year ago about HP's move to get rid of Mark Herd and also defending the acquisition rumor of NetApp which I think I might have started but we were speculating Dave Vellante and I but let's talk about data scientists. So data science is a unique breed. What are you guys finding around the role of data science? You know, there's really two faces of data science. The Ubergeek, I'm a PhD master's in quant shock to a quant shock analytic person who's just a business person. Is there any change in that or? Yeah, I'd say there's actually, so the super geeks, the PhDs out of Cal and Stanford and MIT, the people who are riding the next Hadoop of the world, right? The next, the universities are kicking off those kind of people who are still seated as a Wild Wild West. I was just down in Long Beach at UC Irvine. They're trying to build the next best iteration of Hadoop, right? And so there's still that technology geek stuff going on but I think what's more interesting John is that there is a definite trend to create more data scientist skills and capabilities inside the business organization. The classic organization we're seeing right now. Like you're an example. CMOS. In marketing organizations, we're starting to see more and more marketing organizations hire data scientists or create their own. Take somebody who has the right skill sets who may have had statistical background who may know enough about coding and such but they're really looking to bring those skills in and they're not the kind of people who are going to write new code but the other kind of people who are going to be able to bring in data, be data experts and tease insights out of that data. We're seeing some trends with data scientists that are there favoring MongoDB more than Hadoop because of the ease of use. So we're seeing a shift now where the ease of use of the development side and we're trying to correlate that data as we're going through and through our data team and tease out the insights. For example, is it because they want ease of use? Is it because projects are rapidly developing? Is it because their business drivers are coming from the business teams? All the above, we don't know. What's your take on that? I mean, not necessarily the Hadoop versus Mongo, but the ease of use of deploying, building. As we see the growth of the data side of skills inside the business unit, ease of use is going to become much more important than the ability to write your own. The ability to take what I have and get more out of it because the time to value is more critical. I don't want to go off it to build it and maintain it myself. I want to go and buy it from a vendor who's going to support it and continue to evolve the product. So I think, John, as you see more data scientist skills pop up in the line of business, you're going to see the wild, wild west of Hadoop and open source development given away to products that may only iterate rev every nine months, but are rock solid with support and allow the business people to analyze the data more quickly and get insights out of it without having to worry about building their own. Okay, Bill Schmarzo, my friend, a CUBE alumni been on many times. Always great to have you on from the, you're out in the field, you're in the trenches, you have your own little whiteboard room at EMC and Sunnyvale, you're out talking to customers, CTO, a big data practice for EMC consulting services as an architect and a technical architect, great to have you back again. Okay, we'll be right back with a wrap up of the closing and we're going to stay here. We're going to get Dave Vellante. So Bill, thanks for coming on. That was Bill Schmarzo, good friend of ours at the CUBE, friend also in Palo Alto, new to EMC only for a few years, but Bill's been around the block and what I like about Bill is he understands the business. He goes back and understands the old way, been part of that data warehouse business intelligence mindset and so Bill brings a unique perspective because he's been there, done that and understands the bridge that has to be crossed to get to the new world, the new way. Like I said, you're either on one side of the street or the other. You can't be both, you can't be half pregnant. You are either on the old way or you are on the new way and a clear line is being drawn down the middle of the street and we are starting to see who will cross over to the new way and we will continue to explore that at SiliconANGLE.tv. We are going to be exploring innovation and we are going to be finding the signal from the noise on the new way, whether it's big data or whatever. Okay, so we're going to continue to roll. We got, okay, we've got some SaaS applications up to let me just pull up my little spreadsheet here. So Larry Ellison is going through his keynote. We've been watching a picture and picture of Larry Ellison. Dave Vellante, you're back, babe. Welcome back. You just were in the war room with the team on the horn, kind of like the masters, all the analysts kind of cranking around, huddling. What are you finding? Give us up to speed. Give us a signal from Larry's noise. So I was listening to Larry's keynote, John and he went through basically their Fusion apps update. Very interesting. The two thirds of their Fusion apps are actually being deployed in the Oracle public cloud as SaaS. And then he said, we have lots of customers, lots and lots and lots of customers. And he went through many, many, many examples, trying to make the point that we have lots of customers. So, and then he talked about, he had some classic quotes in here because a lot of people accuse, of course, Oracle, rightly so being, you know, closed and proprietary and lock in. He said, just because we're in the cloud doesn't mean we can't make, we can make everything proprietary. The apps, the platforms, the virtual machine technology, everything is industry standard is what he claimed. Everything, and he's right, everything's written in Java because they own Java. They're using Zen for the virtual machine platform. Of course, they did some work to Zen to make it sort of Oracleized. Just because the app is in the cloud doesn't mean you don't have to interconnect using industry standards. So he really took potshots of people that are saying that they're closed. And then the other thing he point, he made, John, is about social. Talking about his strategy, Oracle strategy for social networking, being that it's not a separate module, it's not a separate four-pay module, that social networking is built into every fusion app. And then he got into the multi-tenancy. You remember a couple of years ago, he poo-pooed multi-tenancy. Basically now he's describing that multi-tenancy implemented the application layer was the only choice that Salesforce and NetSuite had back in the late 90s. Now you can implement multi-tenancy in the database. That's 12C. So he's basically laying forth the pitch for Oracle. As always, very compelling. There's a couple of gotchas in there that he's not sharing with you, like what you're going to pay for this and how you're going to be locked in, et cetera, et cetera. So what's your take on the lock-in? I want to ask that, I've got the Nutanix CTO, DeRayon. They $15 million buy out. People think it's open source because someone's paid for it already. Well, I mean, there's no question that Oracle's got you, and that lock-in stems from the database. I have no doubt that their fusion apps are all written in Java, but the database is really the issue there, John, and it's very difficult to get off the database. Even EMC, we had Sanjay Merchantani on SAP Sapphire talking about is that SAP, talking about how they're moving toward SAP, well, guess what? We heard today from EMC IT that the SAP application is running on an Oracle database. It's just very hard. You can't just rip and replace your Oracle database. You got to bring your company to it's knees. So how many times did he mention Salesforce.com? Several. Certainly more than a handful. He couched it by saying, you know, not to pick on them, but, and then he picked on them. It was very good. You know, classic Larry. So he hasn't really gotten into the hardware wrap yet, but he's really pounding the fusion applications and talking about the underlying infrastructure of the fusion app saying, you can trust us. It's like what Newtonix was saying today. They'll pay a lot of money for trust. That's what Larry's message is all about. He's a genius. Larry Ellison's got the trust message right now. He's just rambling on about exadata analytics, more memory, faster, more reliable. Can we listen? Let's listen to Larry. Let's go to Larry. So many users, some users data in the Oracle. We're going to try to bring the Larry out. It means you're really not going to be using huge amounts of capacity. So it should be cheaper for us to provide and therefore cheaper for you to buy. Axolytics, again, you can see the direction. We have an in-memory database machine. We have an in-memory analytics machine. Oracle is aggressively moving to semiconductor technology for storage. The performance differences are enormous and the reliability implications are enormous. And you're going to see us continually upgrade our systems to take more, to migrate or deemphasize the use of disk drives, mechanical spinning things. And emphasize the use of highly reliable semiconductor things. And of course, we can interconnect all this stuff, not with Fiber Channel, which is very common, not with Ethernet, which is very common, but a much newer and dramatically faster, and guess what, more reliable technology called Infinibend. And Infinibend is much more reliable than Ethernet for a lot of reasons. The software is a lot simpler, for example. But anyway, guaranteed delivery, I'm not going to talk about that. Okay, Oracle Social Relations Management, it's a platform, it's not just a marketing application. So there are really four pieces to the Oracle SRM platform, our social network, which by the way, I demonstrated last year. Social data and insight, the fact there's a lot of data out there. And with the advent of Facebook, not almost a billion Facebook users, more than 100 million people on the Twitterverse, I think that's a word. There's a lot more data out there than there used to be and all of that data properly processed will give you insights about your business, insights about your customers, insights about your products. Social engagement monitoring, I demonstrated that last year, the fact that we can listen to what customers are saying about our products on, say, Twitter or Facebook. And then immediately respond by sending, let's say they're saying such and such a product doesn't work very well or has this problem, you can actually glean that information from the social networks, by processing data from the social networks and immediately respond to people. Either using social media or calling them on the phone. And sending that information to your service department and responding to what people are saying about your products and about your services. If there are problems, you can respond. You discover the problems earlier and you respond to the problems more effectively. But all of these applications, all of these applications are enabled human resources. We help you use LinkedIn to recruit people more effectively. Again, it's not just marketing, it's sales, help you find leads for sales and so on. All of these applications are socially enabled. Again, here are the four pieces of the social universe, our social platform, I mentioned the network. I'm gonna demo, in just a couple of minutes, I'm gonna demo how we process, how you can take vast amounts of social data and process it real time to gain insights about your business. I think it's gonna be really cool. Again, engagement and monitoring and social marketing, again, last year I demonstrated, that we could put up a Facebook presence. I think it was Brookstruct natural groceries that those of you who suffered through that very long demo I did last year. Want you to know, I make this following promise, this demo will be way shorter. People start fleeing already. Here I'm gonna do another demo. This is gonna be a lot shorter demo. But anyway, last year I demonstrated how we could, not only create a presence on Facebook for Brookstructs, we're gonna launch a new product, a new coffee for Brookstruct, and we're gonna orchestrate a lot of tweets and a lot of Facebook posts and all of these things. A lot of people think this is spontaneous, sometimes it is, sometimes it isn't. Sometimes these tweets are what we call sponsored tweets, so I love that expression. What does that mean? Are you gonna tweet this? I'll pay you. Okay, I like that. Better than getting a job. All right. Okay, we demonstrated three of these last year. I demoed the social network, engagement and monitoring and social marketing. This year we're gonna process some big data. All right, we're doing it for time. All right, I'm gonna move quickly through this because I think I want to get to the demo, not torture you. Again, collaboration, the way we interact, each one of our Fusion applications has built-in collaboration using the social metaphor, using the social metaphor. So there are, if you will, threaded conversations, real-time threaded conversations. We can subscribe to different data sources. If you're a salesperson and you wonder if the signed contract has arrived in accounting yet, the social network, the contract arrives, you'll be immediately notified via the social network from the accounting department. Yeah, it's there. You got the deal, and the deal came in in the quarter, so you're gonna get paid next week or next month or whatever exactly it is. So the social network, a lot of the information coming to the social network isn't me posting the data, are you posting the data? It's systems data that's being sent out to keep people informed as to what's going on. One of my key customers, I'm trying to close the deal, one of my key customers has a serious service problem. The service system will automatically let you know there's a problem so you can investigate before you make that sales call. Coming up in the demo, engagement and monitoring. Again, this is the notion there's a problem with one of your products. How do you find out? Monitor Twitter, monitor Facebook. If it's a serious problem and we can figure out what's wrong, you can route that to your service department, you can route that to engineering, fix the product, route it to the service department and respond to those customers who have a complaint or post that you understand that there's a problem with your mapping system and you're gonna fix it. And social marketing, again, you're gonna establish, you can use our tools to establish, quickly establish Facebook presence, presence in Google+, other social networks. You can use the media, again, obviously there are lots and lots of marketing tools that we have that exploit social technology. Okay, lots of cloud customers. Okay, here we go. The demo. I got a question. I might not look like it, but I am the U.S. brand manager for Lexus. I just, I made that career change on Tuesday last week. So I gave Oracle two weeks notice and I got one week to go here and then I'm off to Lexus. They promised to give me an LFA to drive to and from work. Anyway, so the first thing I wanna do, okay, the Olympics just ended and I want to, America did very well, the United States did very well in the Olympics and I want to get everyone driving LFAs. I think it's gonna be a very popular car or some kind of Lexus. It doesn't have to be the LFA. Could be, I'd like everyone to drive a Lexus. I want people to get excited about my brand. I wanna have a big name athlete promote the Lexus. The question is who? Who is the right person to promote the Lexus brand? Now just after the Olympics have ended. And the way what we're gonna do is we're gonna kind of, try to figure that out using, crowdsourcing the information, using the huge amount of data, we picked Twitter. And we really built this application and actually a lot of surprising information came out of it. So what we did is we, over about a 10 day period, we pulled right after the Olympics, we pulled almost five billion tweets from what's called the Twitter fire hose. The Twitter fire hose is a set of APIs whereby you can access the stream of tweets from Twitter. So we used the fire hose APIs, we accessed Twitter and we loaded all of those tweets, five billion of them into an Oracle database. Now to really analyze the tweets, we're gonna have to use a combination of technology, not only the Oracle database for analyzing some of the structured data information, but also Ndeka for analyzing the unstructured information. And we're gonna have to do it simultaneously. It's not good enough to do just structured or just unstructured, we're gonna have to do both. It turns out to answer this question. It sounds like a simple, it's a very simple question, very complicated data processing to get the underlying answer. And we're not gonna do it using batch technology, we're gonna do it all in real time. It's a lot of data. It's actually worse than I said, because a tweet, it's not just 140 characters. There's a lot of structured information in tweet. There are geotags. We know where the person was, if geo's enabled, where the person was when they sent this tweet. We know what time they sent the tweet. We know what kind of device they use when they sent the tweet. We have a lot of information beyond just the 140 characters. We know how many followers this person has. We have a lot of information. So that five billion tweets blows up to a much bigger data set than five billion tweets. We end up with over 25 billion relationships, a lot. 100 million authors, a lot of mentions. So again, a lot of people who are mentioned, a lot of products that are mentioned in the body, that 140 character body of the tweets, obviously there's hashtags and a bunch of other information. Anyway, we know with this enormous database, starting with the tweets, then that thing explodes and it gets much bigger. Well, big data meet big iron. So we have the ideal set, you can applaud if you want to, you don't have to. Very. Big data meet big iron. We have the ideal set of tools to do real-time data processing, real-time analysis of big data. We have, actually, these two machines are now the yesterday's news. We've announced the successors to these machines. But again, we did this right after the Olympics and before the X3 was available, so this is the X2-8 and the Elix, Exolytics X2-4 interconnected with Infiniband. Again, we have all of the Oracle BI Foundation running, tune for in-memory data processing, running in Exolytics. We have the Oracle database, version 11.2 running in the Exadata database machine. Again, interconnected by the world's fastest network. Okay, so let me go, I'm gonna have to move over here. And this is gonna make sense. Can we switch to the computer, please? So, it's very interesting. These are, you can see up here, we've got the five billion tweets and that exploded into a large, this is a dashboard, this is an Oracle Business Intelligence dashboard. Running on, yeah, running on, being produced by Exolytics in real-time. And you can see it's a pretty large dataset, it's big data, the five billion tweets, over 27 billion relationships, you know, almost a billion hashtags, and so on. Here's the first kind of interesting piece of data that surprised me. It didn't surprise me that Apple was the most popular mobile device, or that Android was number two. It did really surprise me that BlackBerry was almost as big as Android, and not that much smaller than Apple. I was really, you know, everyone talking about the end of BlackBerry, but look at this, real data, real data and as a mobile device, not that far behind. Go figure. Be interesting, if I was an investor, I might buy a little bit of that, you know, that rim stock. I'm not making any recommendations, by the way. Let me, let me just do a little legal disclaimer here. I do not own, never have owned any rim stock, and I may or may not be Larry Ellison. This is a, and please ignore everything I say. I think that was kind of in the previous thing, I think it's called safe harbor statement. Okay. All right, so, I mean, kind of interesting, not really the question I was trying to answer, but kind of an interesting, interesting piece of data. So right now we're doing a tweet analysis and we're looking at the data by source, and I'm gonna click on this and I'm gonna change it to topic. And so this is what the tweets are about. The most popular topic was society by a long shot. Second is leisure. I happen to love leisure myself. And third is the Olympics. And you can see the huge spike of activity in this date range around the Olympics. So they're really, the Twitterverse got very, very active during the Olympics. So let's look at basically the top people, the top, I guess the word celebrity is what we use, the top celebrities who are Olympic athletes. And this is an interesting measure because the first measure, we're gonna rank all of the Olympic athletes. Again, we're gonna hire one of these guys, or gals, to promote Lexus. We wanna rank all of the Olympic athletes by followers. And you can see LeBron James is by far the biggest, has lots of followers and Serena Williams, Carmelo Anthony and so on. And I mean, it's kind of interesting, but followers is not the right metric for us. We, just because they have a lot of followers, doesn't mean they're the right, I mean, obviously LeBron James plays in the NBA every year and has accumulated these followers over a long period of time. Again, from the point of view of the Olympics, he may, you know, it may be somewhat different. He has the most followers who invented another idea called reach, looking at just the Olympic range, that 10, 14 day period for the Olympics. And reach is during that 14 day period that followers covers their entire lifetime on Twitter. Reach is the number of tweets, retweets and mentions during the Olympic games. A very different kind of measure. Now again, mentions mean we have to do textual analysis, analyze the instructor data, find how often LeBron is mentioned or whoever is mentioned. During that 10 day period. And how many people retweeted his tweets and so on? Again, during that 10 day period. And let's do that analysis. Okay, first it's by followers. Now let's change this to this reach value I just described. LeBron is still first, but notice that Serena drops from second down to eighth. And Ryan Lochte moves up to second, where Ryan Lochte didn't have close to as many followers. Ryan Lochte has moved from fairly low to fairly high looking at this new measure, which you would think. I mean, obviously Ryan Lochte's, a lot of people never heard of him before this most recent Olympic games. So this is an analysis or a list of the celebrities or athletes by their Olympic reach. By, you know, ranked by the Olympic tweets, retweets and mentions. Now let's go over here to this tab and analyze the audience for these athletes. All right, notice we still have the same athletes over here on the left. LeBron, James, Ryan, Lochte and so on. And interesting, here's a geographic distribution of those five billion or so tweets. Notice how many come from the UK and UK, UK, Scotland, the general area of the UK. I mean, this is very important. Let's say we're a T-shirt company and we're trying and we want to sell Ryan Lochte 200-pound T-shirts, a picture of Ryan in a bathing suit you can wear. And we need to know, we want to do that promotion in London and let people know, we want to send that out to people who are near our store and we probably want to qualify them that they're fairly wealthy and not very smart if they spend 200 pounds for a Ryan Lochte T-shirt. That's a separate query. That's a separate query. That's a separate query. That's a separate analysis of audience, of who to do the promotion to. But it is interesting, again, that we see that you would think the United States would have the most Twitter activity, but around the Olympic time and the Olympic tweets, no, it was actually coming from the UK. Now, analyzing the, that's a geographic analysis, a geographic distribution of the audience. And obviously, we got that where the tweet was, where the tweet device was geo-enabled. You can enable and disable that. But also, we're going to analyze the audience, looking at these guys have lots of interest beyond the Olympics. So the guys who follow these athletes have other interests. And remember, the thing we're interested in is cars. We're trying to find out the name of the person who will be the ideal spokesman for Lexus. All right, so you see down here, you got music, food, so on, football. And finally, we get to cars. So let's drill in on this and rank the celebrities this time by mentions about cars. So people who are interested in cars, just looking at the, let's just reduce the number of tweets to only take the tweets about cars and see how many of them, how they rank these celebrities, or what the reach is for these celebrities. Okay, on cars, things change dramatically. All of a sudden, LeBron has dropped to eighth and Michael Phelps has gone to number one and Gabby Douglas has shown up from nowhere. I mean, Gabby is the, as I'm sure you remember was the Olympic gold medal winner for all in gymnastics for the all around and led our team to the Olympic gymnastics gold medal. And she was way down on the list until we focused on cars. This is an enormously complicated query that where exolytics had to send it back to Exadata to do multi-pass sequel, to go ahead and do this analysis. And we had to add all of you's, both the Oracle database and Ndeka to handle both structured and unstructured information. This is complicated stuff all happening in real time. Most people when they talk about processing big data, they do think about doing it in batch. Really a big map reduce, a big map reduce batch program. In this case, we're using big iron Exadata, exolytics in memory technology and Ndeka the Oracle database to process unstructured and structured data in memory in real time. Okay. We're getting very close. We got one last query. We wanna do this by brand. Now it's not just gonna be cars in general, but we're gonna break it up by brand. And here's the reference analysis. Oops. How do I back? Okay. All right. Here is the reference analysis. Now this is very interesting. Okay, so we got the answer to our question. The answer to our question is Gabby Douglas. And then come back to Gabby with the second choice of Michael Phelps. You know, I would really, if I'm Toyota, well, before I make that comment, let me just go down and describe this screen. The, if we look at the different hashtag groups, different communities that are interested in Lexus. And we've done sentiment analysis on the tweets. That is, if people said, I hate Lexus or my Lexus burst into flames and they tweeted that, we consider that a red or negative tweet. I'm not sure, you know, or the insurance on my LFA costs more than my house. That would be a negative tweet. So we've actually used our, you know, and Deca technology to look at the tweet and establish, is it a positive comment? Is it a neutral comment? Or is it a negative comment? And these groups, we see here, we only have one negative group you can see and that's on insurance. That's popular. They're playing for you. If he showed you more of it, there's a lot more negative stuff. But the neutral, the neutral tweets are in gray and they're not used in the calculation other than say how often they occur. The positive tweets are in green and the negative tweets are in red. So, just a general comment about cars, it's considered neutral. Mention of the Toyota brand is considered positive and that mention of the LFA supercar is considered positive. And a mention of the Nürburgring, the N24, the Nürburgring 24 hour race is considered a positive. And lots of people who talk about Lexus talk about the N24. This gives me as a brand manager, I know we're thinking of entering an LFA in that race next year. A lot of people who are interested in Lexus follow that race closely. So I get the insight, that's not a bad place to spend money. We should have an entry in that race. We should do promotions around that race. But again, this is analyzing big data to get insights about our customers and our potential customers and what's the best way to promote our brands to those customers. And finally, the right spokesman, and I'm very excited about this, is the LFA brand, or excuse me, as the Lexus brand manager. The right spokesperson for this is clearly Gabby Douglas. And I'm definitely gonna offer her a lot of money for this deal because Michael Phelps is six foot five and Gabby Douglas isn't even five feet tall and so she's gonna make those Toyotas, those Lexuses look really big and luxurious. So I want her in my commercials, not this big swimmer. So we're gonna make her a big offer. So again, you can look at this. This was a very simple question that required an enormous amount of sophisticated data processing to get the answer to. But we can now get the answer. We now have access to this data about what people say about our products, what people say about our services. We can now analyze that and get said to sentiment analysis, figure out what topics are important to these people and figure out who are the most influential spokespeople. This is something that we've just had to guess at in the past. This is sophisticated technology, processing big data. And that's it. Thank you very much. All right, we're back. We're back at Oracle Open World. John, we just saw Larry. Yeah, I mean, I got to say I'm geeking out on that his presentation, loved it. I saw the Twitter demo at the end was fantastic. You know, Larry's working knowledge of big data is instilled in kindergarten. You know, MapReduce is not just batch. He's trying to pigeonhole Hadoop, but that's okay, the demo was smashing. That is an actual real life example in my opinion of the business value of big data. The great example, it shows you the power of big data. It shows you the power of, you know, what big data analytics can do for you and the requirements behind it. Now I'm not saying that Oracle is the only bridge to get there, absolutely not. We had MongoDB on earlier. Oracle, if you're buying Oracle, it might be a great solution for you. I wouldn't pay extra for that because you can get it somewhere else with Hadoop or Mongo. But great demo, I loved it. Loved it, loved the demo. It was a great demo and of course, John Oracle would love for you to believe that Hadoop is batch and Oracle is real time and I've just got one thing to say to that, which is, big data meets big iron. As you said, there's other ways to skin that, Cat. Isn't that, isn't that EMC's logo? Oh, that's cloud meets big data. Big data meets big iron, yeah, that's good. But I agree, the demo was substantive and very complex, as Larry said. Using a lot of data, so is Oracle big data? Oracle is now cloud meets big data. So EMC, Jeremy Burton was on earlier. It's like co-opting your messaging. So my prediction came true. I was thinking that it might not have been the case, but Larry is the king of big data. As of right now, Larry Ellison is the king of big data. And he just proved it again. And Joe Tucci's got to stay on hard and get it going, so you know. So today's keynote, John, was about apps. Spent a lot of time on fusion apps. Spent a lot of time on talking about how social is a fundamental component of the fusion apps. Talked about a big data example. And very importantly, I thought, brought in in DECA as a semantic platform on which you can leverage and build applications. And DECA was a great dream deal for them. I don't think they've overpaid for that deal either. They paid big money. I mean, you know, North of a billion, right? They had solid install base accounts. They had the whole structured data within the enterprise nailed. And DECA was not a bad technology. Again, that's proof points for DECA right there. A lot of people wondered why. Well, a lot of people believe that some kind of search or semantic app or apps will be built on semantic layers. I mean, here's the beautiful thing about Larry Ellison's orchestration of his keynote. One, he goes after Salesforce heavily, but the way he positions his social media technologies is different than, say, Salesforce.com. Salesforce.com, no offense to Benny off, you know, your people will probably smash me on Twitter. But you know what? You're kind of piecing in disparate technologies around your social media. And that's a Cisco strategy. It buys you instant pop, a big splash in the pool, but you got to put that all together. So, you know, where Salesforce possibly could fall down is their mashup of these disparate companies that spare technologies. That doesn't play well. Larry Ellison takes much more of a technology buildout approach. So it's clearly he's stating here is that Oracle is going to enable social as a fabric across every single piece of Oracle technology, not try to jam it together. And I got to tell you, that's what Salesforce.com is doing. Salesforce.com is jamming in different disparate components and then as a sub-optimized situation for customers. Well, so John, as I say, you know, today was about the applications. Sunday night was about hardware. So you had the software company talking about hardware. This is what Chairman McBurton said today, you know, really focusing on the application space. As I said before, the other thing he said earlier in his talk was that two thirds of the fusion apps are being deployed as SaaS. So what are Oracle's, you know, what's the perceived weak spots? Cloud? For whose weak spot? Oracle. Cloud, SaaS, or late to cloud? Salesforce, hurting them in SaaS. Hardware, sole, sun, hardware. And really focusing on flash to boost up the hardware perception. Talking about how much demand there is for SaaS. Focusing on trust and security. The thing that CIOs are most concerned about. As always, very, very crisp messaging by Oracle. Oracle is just very tight. I mean, I was talking about the social, the demo with Twitter was kickass. I thought that was fantastic. Very good presentation. I'm going to give them an A plus on that. And like you said, Oracle is competitive. We had many ex-Oracle executives on theCUBE today. Jeremy Burton and Dheeraj from Newtonics. And they all say the same thing. The culture is very specific. Check the boxes. We talk about this every year. Oracle, we do kick them down. We kick on them a bunch of times. We do needle them because Larry's style, we like to emulate that back at them. We don't mind doing that, but we got to give them props, Dave. We got to give Oracle props. They're tight, they're aggressive and they're competitive. They are the blanket of comfort for their clients. Because the switching costs for an Oracle customer are massive. So what they do is they create a comfortable blanket to make those switching costs even harder. And that's service and that's treating them like royalty, having big blowout parties, buying yachts, doing the things that Oracle does to service their high-end clients. We asked Jeremy Burton on theCUBE what he would do differently if he were running Oracle. And he really didn't say that he would do anything differently. He said, well, if I were in that position, you know, he looked at them and said, well, very, very successful financially and strategically obviously doing very well. So if you were in that position, would you do anything differently than what Oracle's doing now? Would you go open everything up? Would you give away the franchise, go open source, commoditize your value of opposition, or would you act like a monopolist? What would you do? Would you act like a monopolist? He said the performance has been proven that it works. So I don't think he said that directly, but we know what he was saying. He would, you know. Yeah, I'm paraphrasing, but Jeremy basically said, hey, I'm looking at them and they're very successful. He didn't say I wouldn't change the thing, but he didn't articulate anything that he would change, right? EMC obviously is coming at it and VMware from a different angle. As is Salesforce. You know, here's the thing, John. Salesforce, he said today earlier, NetSuite was actually the first company to start with SaaS. Isn't that NetSuite? Isn't the investor in NetSuite? Yeah, I don't think so. I don't know. But anyway, let's face it, Salesforce really started that trend. And then obviously Amazon with AWS. Some good lines in there. Ignore everything I say is what he said as jokingly. And he also mentioned RIM as having install base close to Apple and Android. He said, if as an investor I'd buy RIM and then he went off this disclaimer. There was some good comedy moments by Larry Ellison. I got to say, he does make me laugh. He's very entertaining. He is the king of the industry, captain of the industry. And you got to give him props for that. And I got to, you know, I respect the guy for that. And, you know, he's competitive. You know, he can take credit for a lot of stuff, but that's him doing his thing. All right, John, well, so we're here. This is day two. We're back tomorrow. Not as heavy a day tomorrow, but we got, you know, more on networking, more on Flash, more on the cloud, more on big data. Bring it to you live. So we should reset here. This is theCUBE. This is SiliconANGLE's continuous coverage. This is our third Oracle open world. SiliconANGLE.com for all the news and breaking analysis. Wikibon.org is where you go for all the deep research. Go to SiliconANGLE.tv. And also check out our playlist on YouTube.com slash SiliconANGLE where you'll see all the events that theCUBE has been to. And we introduced today our new research product called Trend Connect, SiliconANGLE Wikibon, our first joint research project where we go out and we survey the landscape and we put a panel together and we introduce that Trend Connect. You go to slideshare.com slash SiliconANGLE and you'll see a copy of a sample report that we provide to our clients through our research team at Wikibon. And if you become a client at Wikibon, you get that free sample report of any vertical, any market you want. And it's more in depth and you have more data behind it. That is a four, there's a fee-based service for that big data report. We will be putting summary reports like that out there for free if you want to digest it for free. Go to slideshare.com slash SiliconANGLE. And of course go to SiliconANGLE.com for all the action, we're not missing the story. And also for the breaking analysis, something new from us. We're going to provide really hard-hitting analysis fast. The fastest, highest quality analysis on the market. It's not breaking news, breaking analysis. And then go to Wikibon.org, that's the site where you can go inside the ropes and go deep dive and post your own stuff. Comment on David Fleurer's research. It's a production environment where it's a community-based research site, Wikibon.org. And of course, that's all going to be coming together the next few months on SiliconANGLE.com and SiliconANGLE.tv. And this is theCUBE, our flagship program. We go out to the events, extract the signal from the noise. This is where the action's at, that's where we'll go. We'll see you tomorrow. Day three of Oracle Open World, blanket coverage. I'm John Fleurer with Dave Vellante. See you tomorrow.