 Live from Las Vegas, Nevada. Extracting the signal from the noise. It's theCUBE, covering Informatica World 2015. Brought to you by Informatica World. Welcome back everyone. We are here live in Las Vegas for Informatica World 2015. This is theCUBE, our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. George McCose, the segment, George Gilbert. Who's getting wired up right now, but our guest here is Ivan Chong. EVP and Chief Strategy Officer for Informatica. Welcome to theCUBE. Oh, thank you for having me, excited to be here. We love talking strategy with the Chief Strategy Officer, EVP, you got the chess board in your office, you got the lay of the land, all the executives around the table, product groups are kicking butt. What's the strategy? I mean, I'll say you guys are going private. We saw that news, and that's still not finalized, but Michael Dell took his company private. He's so happy, smiling. He's a spring in his stab, he's doing great. The big data world is certainly got a lot of growth. Well, I think just one way to think about the strategy is that we love disruption. We love disruption because we think the world is coming to realize the value of data, right? And as you mentioned, data is having a life of its own outside of applications, outside of systems. And it's going so diverse that having an integration platform, having a data management technology is becoming much more valued by organizations. You guys have an engine, explain what's going on at the engine. So you have an engine, what's the core engine for Informatica and then what's wrapped around it? So we like to talk about an intelligent data platform and it starts with the ability to get at data wherever it exists, right? Not just data because it sits in an application system or not just data that's part of a business intelligence system, but anytime you can harness data, we want to get at it. So those are the connectivity feeds that we pump into the engine. The core engine we call Vibe, right? It's got this ability to process data no matter where it came from. It uses metadata, so it retains context. It does a lot of processing with regards to transformations and aggregations. It applies data quality to be able to produce more valuable and trustworthy data. And then on top of that, we add the metadata to draw context and that's where you'll hear about a lot of our exciting new announcements with regards to security, with regards to discovery of data through live data map, with regards to data preparation. And then on top of that, we do provisioning, not just to systems like business intelligence systems and database systems or Hadoop systems, but to individuals, business people that are using data to be able to make decisions and to harness it to be able to gain insights on their business. You said something interesting about all the processing that goes on in Vibe and then the context that gets added from sources. Traditionally, when you're processing, at least near real time, stream processing, adding the context can slow it down. How do you manage that trade off? I think what we usually do is we provide multimodal processing, right? So you can process it inline. You can start to process it in batch or offline. And so we allow the customer to trade off. How critical is it to have that sort of sub-second response? How critical is it to have it after you've closed out your quarter? So that's why you'll see us investing in all sorts of low-latency technology as well as batch processing technology. So you'll see us in stream processing as well as Hadoop. Okay, so the stream could be like a tunable latency ingest into Hadoop where you do the rich context, the historical context, or it could feed something near real time or real time. Yes, yes. We have to create a platform that supports both modes of operation because based on our legacy and heritage in analytics, we see the value of both and we understand the technology requirements needed to support both. So about real time, this is really interesting. And in context, the cloud. So we talked to Ash earlier about the cloud business doing really great and the consumption that customers are saying is I'm going to move to the cloud, eventually hybrid cloud's certainly hot right now. But real time, I want to get my resources in whether it's content, sales people, assets, to my customer in real time at the moment of impact or at the moment of relevance. It's a very hard thing to do. How do you guys look at that real time with the cloud architecture from a software standpoint, from a data standpoint? Because it's a combination of things, you got to have the data. You got to understand the data. You got to have low latency transit to the endpoint. What's the, how do you guys look at that? Well, we look at it in a number of ways. One is through our platform, right? So we talked about having the different modes of ingest, right, so there's stream processing as well as batch processing. The other way to look at the problem that you've talked about is you need to be able to sense information in real time. You can't just wait until transactions like orders or sales are closed. You need to be able to detect customer intent, customer sentiment, right? So we talk about the platform, our intelligent data platform having the ability to be able to infer things through machine learning, to be able to do natural language processing so that you can take any amount of data and then you can glean insights from it that allow you to make those real time decisions. Another interesting thing you said, applying all these different analytic techniques. One of the things that we've seen with the rise of sort of Hadoop based platforms, even though the uses for most enterprises are primitive, they're unbundling what was a uniform kind of data platform or an integrated analytic pipeline. So we have machine learning, we have the stream processing, we have graph processing, all the different modules that are growing up. And it's a little bit hairy. What's, how does, how do you serve up your components to keep them integrated and low cost of ownership? So what we've done with our platform is we've been able to keep a consistent way of tracking the metadata. And the guiding principle behind that is that the metadata should be able to function regardless of where the data came from. So if you started by processing data that came out of a mainframe, and we've had customers that do this, as you adopt newer forms of data storage technology, maybe you go to a relational database or maybe you go to Hadoop, the same rules that you use to process that data apply in the new context. And that's the guiding principle behind our metadata. Spark is coming out from a different sort of corner of the universe where they have this sort of single execution engine and little wrappers around it where each mode of processing can sort of chain into each other mode of processing. They're not discrete engines. How, how do you compare against that? Well we see Spark as a platform that we adapt to. Just like we have customers that are storing data in HDFS and Hadoop, that stack is going to encompass Spark. This morning we talked with some folks that were experts from University of Wisconsin and that are part of Berkeley's AMP Lab that delivered Spark technology. We think it's just another engine that we can harness so that we can make use of additional technology for our customers. So when we talked about, you asked about the stack earlier, we were very careful to separate the metadata in the context from the engine technology. So the mere value add across all these different types of pipelines is the metadata. Yes. All the way through the source. Yes, that's a secret sauce. Yes. And we can do more things to innovate by gleaning more information about the metadata. So just being able to trace the proliferation of data for security purposes, that's one example. That's huge right there. That's one example. Because now you're tracking, it's geo data, geo tracking, geo space, whatever you want to call it, you can look at data out in the hinterlands of wherever, cloud, device, and understand context of data. Yes. So that's a security notification beacon, whatever you want to call it. Yeah, specifically the context in that case is the sensitivity of the data, right? Because people are drowning in data. So you really want to focus in on what is the data that I am obligated to protect and to secure? And so we have the ability to detect what is sensitive data? What is the date of birth? What is the social security number? What's the credit card number? And then we highlight that data as it flows through the various systems that our customers have. And then we identify, here are the things that you need to focus on to be able to protect data. So your competitive advantage then, if I was just pretending I was working with you guys and say, hey, okay, we've got the metadata secret sauce, let's lay around machine learning and NLP and other cool algorithms to surface notifications around what's important in real time. That way that- Metadata has always been the strength of our company because to be Switzerland, right, to be able to allow for customers to have a very heterogeneous compute environment, you need to be able to focus on the metadata. Can you give us, in addition to security, some more examples of flowing that metadata from the source all the way to the target across different engines? So this new product that we announced last year, REV, it's getting huge adoption, and the premise of it is to allow everyday business users to do what we empowered IT developers to do, which is to be able to blend data together, to shape the data, in case it has different cardinalities, to be able to prep the data on their own without having to go back and forth with IT. Now, to do that, we need to have our metadata. We need to understand, through our architecture, what the business user is doing to the data, right? So when they're blending the data together, it's through suggestions on the UI, but underneath the example you asked for is we're capturing metadata, say, we use the join key on a specific set of fields. It's almost like you've got a log of activity across all the transformations of the data. That's right. In the REV data preparation product, we call it a recipe, right? So that's what the business user sees, but underneath its metadata, that then can be understood and ingested by our traditional engines, either in the cloud or with PowerCenter. So this sounds like one of the differentiators that Cladera hangs its hat on is the governance capability. But this sounds like it would be working across a perhaps wider array of analytic engines and data sources and data targets. Yeah, so Cladera's a great partner with, we work with them because we share very common objectives for our customers. And I think they are a great solution for governance. The way we fit in with them is that they focus a lot on the persistence layer, right? So they've got a very, very valuable economic proposition for storing data, large volumes of data. They've got good technology on top to be able to do the ingest and management of the data. And so when we provide our metadata, when we provide our architecture on top to be able to help them do discovery and cataloging of all the data assets that they've got, you get some of the demos that we showed this morning which show you how you can start with lots of data stored within a Hadoop system and then automatic ability to catalog the information search and discover information and then prepare it. Okay, we're getting the hook here but I want to ask you a final question. What M&A's on the horizon? What kind of white spaces are you guys going to be filling in with M&A? So, perhaps taking over Oracle since the data's more important now than the data manager. Well, I think the M&A strategy that we had before our partnership with Permura and CPPIB is still very much in play, right? So we invest heavily in all data related technologies. All right, so MDM is a big area for us. We continue to invest in that area. Data as a service, cloud technologies, all those cloud integration technologies that can increase the value of how you can manage cloud assets from software as a service and infrastructure as a service. All those are things that we actively discuss and pursue. Well, thanks so much for sharing the strategy and the input, George, hitting you some hard questions that you handed them like a pro. Thanks for coming on theCUBE, really appreciate it. Well, it's been my pleasure, thank you. Thank you, we'll be right back. This is theCUBE, we'll be wrapping up in the short, right after the short break.