 Okay, welcome back to theCUBE. This is SiliconANGLE and Wikibon's exclusive coverage of HP Vertica Conference, HP Big Data 2013 is the hashtag on Twitter. I'm John Furrier, the founder of SiliconANGLE. I'm Joe with my co-host. Hi everybody, I'm Dave Vellante at wikibon.org. Lawrence Schwartz is here. He's the vice president of marketing at Attunity. Lawrence, we know you, of course, from Tokentech. Welcome to theCUBE. Good to see you again. Glad to be back here. Yeah, so it's an impressive event. When I first heard that HP was going to have an event in August and it's the first user conference, I wasn't sure, I said maybe they'll get 100 or so people here, but 500 users and 700 people, it's pretty good. So anyway, again, welcome. Tell us what you're doing at Attunity and what you guys are all about. Sure, sure, no, glad to be here. And yeah, we're very happy with the turnout as well. Yeah, awesome. We've gotten to work with them over the last couple of months and gotten to know them well. For what we do at Attunity, we've actually been around for a while. We've a company that got our history and changed data capture and working with that on a lot of different data types and data systems. So we were doing stuff with HP and non-stop, for example, for a decade or more now. So we know the guys well. And what we've been focusing on as of late is really how you get real-time data integration going. So when you want to have multiple data sources coming into a data warehouse, how do you move the data in there efficiently, quickly, without a lot of development effort. And if you want to do real-time BI and really look at offloading queries and things like that, those are the things that we can really help with simplify the whole process. So what we're here to announce is our partnership with Vertica, that's part one of it, and part two is having a product that supports the vertical offering and being able to do this. We do this for many other types of data warehouses, but we're very excited to be able to do this for Vertica as well. So I wonder if you could talk specifically about how data is changing. Everybody talks about unstructured data. 90% of the world's data is unstructured, but I wonder if we could sort of double click on that a little bit. How specifically is data changing and how is the industry generally in attunity specifically taking advantage of that? Sure, sure. Well, I think what you find is that there's data residing in many different places already existing within on-premises and moving it around there is a key question. The different data types that are now available, as you mentioned, are structured and unstructured. There's different types of content as well. You're talking about social, you're talking about HR records, you're talking about traditional data stores. So how do you move those all in to a place where you can do sufficient analysis on it, have the horsepower to do that? And that's what we support. So you've got Vertica as a fantastic platform, for example, for doing all of those analytics. And then, of course, everyone talks about the need for data scientists on top of that to run it. But what I think it's less talked about is how do you get the data in there in a timely fashion? How do you manage it from multiple sources? And not just the traditional ETL process, which a lot of people are familiar with, Exchange, Transact, move it and load it and do that type of work. But also, how do you move it more efficiently, do more of an ELT type of process? And we can do that very effectively. So when you talk to customers, who owns the data strategy? You know, you hear this talk about a data czar, we were at MIT a few weeks ago, and we had a discussion about the chief data officer. Do you see organizations thinking about that, or is it more technology led today? Sure, sure. No, I think it's very much driven by the users of the data, and what they want to do with it. So again, there's that kind of need out there to get somebody to manage all different types of sources, all different types of data and kind of synthesize it together. So those, I think, are driving the real demands now, not just the IT departments that had traditionally to take care of internal records, right? So they're driving those external demands, and that's what's driven, I think, a lot of the growth in platforms like Vertica for doing the analysis. And then what's helped us grow underneath that is managing all those different types of data coming in. So yeah, there are different drivers today than there were a few years ago. So do you think a role will emerge where there is sort of an individual or a group that is responsible for the data architecture, the data direction, the data strategy, or do you think that will be more distributed? You know, maybe, obviously it'll be both models, but what are you seeing? Is there any, do you have any evidence of any kind of trends emerging? And what do you advise people? Sure, sure. I think the bigger organizations out there will develop those, you know, companies in season house and bring in the people internally. I actually see for smaller companies, there are some very creative companies out there that'll help make that whole analysis process a lot easier, right? Take, for example, you know, some of the big CRM tools out there that are in the cloud and everybody uses. They can be very challenging for the traditional SMB, you know, 50 to 100 people figure out how to use that data and make the best of it. And so I think there are a lot of external agencies that can kind of come in and help do that type of work, you know, with their products, with some consulting. And so that might not be in-house for a smaller place, but there are some great tools developing for that, that side, as well as the bigger places with the in-house technology. So you always heard a lot about, you know, the single version of the truth that was sort of the promise of a lot of the data warehouse and business intelligence systems that have been put in place in the last decade. And then, you know, this dupe thing comes along and people say, oh, we're further from a single version of the truth than we've ever been. What's your take on that and how are you and your clients wrestling with that problem? Yeah, yeah. So I think, you know, Hadoop and some of the unstructured capabilities out there get to a fundamental problem, which is a lot of people don't know the questions that they want to ask, you know, a priori, right? You know, it's not necessarily setting up, you know, no offense to the name theCUBE, but people, that's the honest, I want to start with theCUBE, you know, for the data warehouse. They just want to have access to the data and be able to ask the right questions later on. CUBE's good for video, but it's not always great for data. Exactly, there we go, that's the way to think about it. But, you know, once you have the power to really do, you know, move the data to where you need it to get to, have it on a powerful analytical platform and then be able to ask the questions, then be able to do the transformations and whatnot, those are different ways to think about it and offer a lot more flexibility than just the traditional model. So I think that's what allows people to do is to deal with the questions as they think of them and sort through the data, whereas you don't have to think about it as much beforehand and it gives you a lot more flexibility. Lawrence, talk about some of the use cases, Colin was on here, obviously it's a user conference here, what are some of the use cases that you're seeing with your customers and why the partnership with Vertica and why is that important? Yeah, so I think when Colin talked this morning in his keynote, he talked about the different pillars of the technology and what's important. One key aspect of that is, again, how do you get the data there and move the data over, right? So that's the leg that we're really helping Vertica out and I think what makes it a very interesting, symbiotic relationship, which is they've got a powerful platform, we've got the capabilities to move it there. So typical use cases are if you look at somebody that's got traditionally lots of different sources, they're trying to do real time analysis on it so they're kind of offloaded from some of the other data warehouses that they have, do the queries kind of maybe real time in a Vertica type system. So that's one area, the other part is just straight loading. So I can give you some examples that we've had with other data warehouses in general because of Vertica is just a new partnership for us. But in general, we work with places like on the finance side, you take like Equifax, right? And they have all the consumer credit data that they're looking at. And again, they're pulling it in from multiple data sources. The key advantage for them is they want to be able to look at it and do analysis in near real time. And that's something that you can't do with an ETL process, right? Just move it over, batch it, get it over there overnight, right? If you want that in near real time, we see that happening in manufacturing as well. So when you look at, we have another company, Smart Marginal, in the manufacturing space. And what they do is they're pulling in all their distribution and another information on the manufacturing process from service from all over the world. And then they want to do that near real time analysis of it and what's going on. And getting the capability to pull in the data from all those sources is critical. Talk about some of the BI challenges out there. So you have old way that has software inherent software that's based on slower compute, slower access methods. And so that's old. Now you have Vertica and other folks enabling basically low latency, massive speed, trillions of rows, billions of rows. So that goes away. That's going to change the software architecture. What do customers need to know about the different approaches to BI? There's companies like Plattfora, that Looker, you guys. And there's a variety of solutions. What should they be thinking about? What are the core questions that they should ask themselves when selecting a vendor in the BI space? Yeah, yeah, well it's a great question because there's definitely a lot of different stacks to the BI space. And we're very focused on getting the information in there so we work with all those types. I think being able to analyze the different types of data sources quickly is very important. I think what's also critical to this is a lot of time with these more powerful platforms, you got to start thinking about what you can do with that power that you have in the platform, right? So one thing that you have the capability to do now is get the data in there as quickly as possible. And then you could do your transformations and other things as you need to do them later. So that's kind of one thing that you could then do and then that gives you the flexibility to ask questions as you need to or as they come up. The other thing that you need to think about or the customers need to think about with any kind of BI stack that they choose is, it's great if you can find the data scientists these days to actually fill in that slot. It's great that you have the platform. But when you want to pull in a new data stream or pull in new sources, if you can't do that flexibly, then all this computational power you have, all this brain power, all the data scientists you have, it ties their hands and limits their ability to do things. So you really got to find a full stack. So diversity of sources and then as one issue. And then there's also new sources on top of that that come in in real time. Yeah, exactly. It's the new sources and actually getting a new source up and doing all the scripting that you might do with a traditional ETL. That can actually take months of developer time. It's very, it's not very well automated. So you want to find something that can automate that whole stack. So again, it's that time to value as quickly as possible with the different information sources. And it's not like a geeky coder. It's like a normal user. That's the key in the front end. That's the key thing. And that's what we especially help with. A lot of the setup for moving over databases can be complex, right? You're moving over schema. You're trying to change things and try to set up. But we actually make that a, we call it a click to load interface. And it's literally a few clicks on a GUI interface. Because again, we want to make that as simple of a process of going from different types of databases into where you want to get it so you can actually do the analysis. So what's your core innovation that you guys are enabling with your product? What's the core innovation? Absolutely. So there's a couple of things that really drive what we do. One is that simplified interface, which you might ask, well, anybody can do a simplified interface, right? But what really differentiates what we do is we've been doing this for well over a decade working with all different types of data sources, everything from the oracles to SQL servers out all the way back to mainframes, right? So being able to do those transformations and understand those interfaces and being able to speak all those languages as we will from different databases, that's kind of some of our core IP. And being able to smooth that and make that behind the scenes so you don't have to worry about it. The other thing is the speed and performance, right? So when you look at, for example, streaming data over and moving it from one place to another, people traditionally use change data capture to do that. But that's more optimized for most databases today around a transactional system because we're talking about data warehouses, we actually do a very efficient kind of packing of information before it comes over. So every time you send something over, you can actually get more efficiency out of it and get it near real time. The last thing that we do that's very different from what you might see in an ETL solution is when you go from database A to database B, instead of having to kind of touch down and move everything and transform it and move it over, we keep that as lightweight as possible. In some cases, just doing complete in-memory streaming. And again, that's much faster, less equipment that you have to buy and set up. And again, simplifying that whole process. So why the partnership with Vertica then? What was the big aha with Vertica? Yeah, yeah. So I think for us, we want to be able to work with the best in class solutions out there. We look at customers today, when they look at their data environments, they're really picking and choosing the best solutions that they can for different parts of the databases, the data warehouses they do, and we see Vertica coming up more and more. So we want to be part of that from our end. And then I go back to the value that we really deliver for Vertica, which is again, simplifying that process, not making it too developer centric and getting it into the Vertica databases or data warehouses quickly as possible. I think that's a high value for their organization. So really, it's very useful for both of us. How about data like video data? Or even metadata, that's either specific to video or maybe we have a general metadata discussion. How are customers looking at solving that problem? I mean, it's a hard one, right? You can transcript the video and then you've got the words, but is there anything on the horizon that you see that is going to allow people to conduct more analysis? You've seen a lot of video surveillance. It's sort of the new wave. I think we could talk about that a bit. Yeah, well, it's interesting. From where we see video from our end, and this is actually another area that we've seen a lot of interesting growth for us, is people are using the cloud a lot for their video assets. So if you have organizations that are spread worldwide and they wanna manage the video transfer and how they move it around, the cloud's becoming more popular. So for example, we'll work with Amazon S3. We have a very big partnership with Amazon Redshift. So all those things that I mentioned that are hard to do on-premise, moving those files, and we work with a good number of those customers for us are in the media and entertainment space, it's moving those large files into the cloud. Now what becomes a bottleneck right there, right, is that whole, how do you make the best use of the links in the bandwidth? And our technology's been optimized to do that on-premise and we've optimized to do that for the cloud as well. So that's at least our picture of the video world and where we see those pieces moving. What's the nature of the Amazon and generally in the Redshift, specifically relationship? Sure, sure. So we are a partner of theirs. We've been invited to a lot of the webinars they've given recently. We have a lot of joint customers and particularly, we have it on S3, we have it in multiple areas for the cloud, but particularly when you look at the Redshift side, which is basically their data warehouse in the cloud, they've looked at how do we make that more performant? How do we get data in there more quickly? So I'll give you one example. One of our customers is domain holdings and what they do is they do all that online registration of domains and they try to do the optioning and all these other real-time processes they want to run. So that goes back to our real-time need and real-time effort and capability. The question that they had is, okay, I've got lots of different data sources. It's a big amount of data. They looked at a project that was going to take them three months to move it all over. With us, they were able to do it in minutes because we're that simple and then getting it all set up and going and that's the value we bring for Redshift in the cloud, so awesome. All right, Lawrence, we'll listen. I really appreciate you coming on. We're going wall-to-wall here, John. Bruce Yen is up next from, yes? Good, we'll get to see Bruce again. We have a good comfortable experience with Bruce. Absolutely. It'll be great. We'll be right back after this short break, which is exclusive two-day coverage of HP Vertica's user conference. The hashtag is HP Big Data 2013. We're watching that. So if you have any questions, hit the hashtag. We'll be right back after this short break.