 Dave Vellante here with Jeff Kelly. This is ReInvent 2013, and you're watching theCUBE. SiliconANGLE's live production. We go out to the events. We drop in our mobile studio, we find the best guests, and we bring them to you, extract the signal from the noise, share with you our audience. We appreciate all the tweets that you've been sharing with us this week, all the questions that have been coming in. John Rooney's here, he's the director of developer marketing at Splunk. John, this is like heaven for you, because we're just surrounded by developers. Yeah, now it's a fantastic show. We were here last year, and we were just completely blown away by not only the number of people here, the quality of folks here, a lot of our partners are here, and it really feels like a huge center of gravity in the space. I mean, the development community has, it will coalesce around major initiatives. You're certainly seeing that with OpenStack now, a lot of enthusiasm there, but Amazon really is the icebreaker here. Talk about the developer community, how Splunk looks at the developer community. Yeah, I mean, I think we've realized, like a lot of other folks have realized, that developers really are key stakeholders, not only for sort of classic technology selection, but sort of driving that technology forward. And Cloud, for example, a lot of the early use cases, and I'm sure AWS would say the same thing, was developers looking to spin up environments without working through traditional IT channels, and that ease of use, and that quick time to value is a big part of it. And from a Splunk standpoint, we look at developers as being huge enablers of sort of driving the value that Splunk delivers in terms of providing operational intelligence in the data across the enterprise and in different ways. And that could be integrating insights from data into other systems within the organization, so integrating it into a line of business applications, building applications on top of it. You know, really looking at and also looking at how to utilize operational intelligence across the entire product development life cycle and things like DevOps and continuous integration and continuous deployment, which are all, again, sort of tip of the spear. So we've been having crowd chats all week, it's our application that we use to interact with our audience. We've been talking about the, you know, Amazon turning the data center into an API. We talk about the API economy, the data economy, your shirts as rest, you know, all caps, restfully APIs for the Wicked. What does that mean? Well, I think it's really, we just wanted to make sure that, you know, Splunk is a platform and we really want to invest in that. We think about that as a platform. So everything, if you are familiar with Splunk, you've seen Splunk, a lot of times you focus on, oh, what does the UI look like in the drag and drop and the dashboards, but there's a REST API layer that sits on top of the engine that developers can programmatically do everything that you can do in the API via that, or via, in the UI via that API. So we really want to make sure that it's an open platform, it's an extensible platform, it's extensible in terms of getting data in, it's extensible in terms of getting data out, it's extensible in terms of building applications on top of it. And we really wanted to, you know, the RESTful part of it is we want to, you know, bring our technology to where developers live. So web's open standards like REST and ensuring that we have support for popular languages, your JavaScripts, your Java, your PHP, your Ruby's, and that we sort of address the polyglot of developer tools and languages out there. So talk about AWS, what the relationship is, you're part of the ecosystem, where do you fit and what are you guys doing together? So we're doing a lot of different things. The, yesterday we are one of the partners for the cloud trails announcement, which is a huge, I think a huge value for both our customers and AWS customers. Yeah, describe cloud trails for the audience. Sure, it is, basically it is a mechanism by which AWS customers can derive log information from popular AWS services, so they can view access control utilization. Like what are people actually doing? Performance, we're talking about performance this morning, yeah. So, and what we, our integration point is, and it's a perfect use case for Splunk and the Splunk platform, is we ingest that data and then you're able to visualize there's a Splunk for AWS app that's available now via the Splunk app marketplace that enables you to see not only usage data, performance data, billing data, so what is the utilization of reserved instances to give folks that sort of full 360 operational view into what are your investments in AWS? And I think it's a huge step forward. I think it provides a lot of folks, a lot of the visibility and access into what's going on into my, into sort of my little section of the cloud that I'm using on AWS in a way that maybe they were used to in a sort of a non-prem world. So you're going to help people save money. You saw the slide Andy put up the other day, 143 million saved and proactively Amazon shooting on email. So hey, if you did this, you could save a bunch of dough. We want to arm our customers with as much information to sort of, yeah, make the right investments on their part. And then in your key part of that, you know, to me, that's so brilliant, Jeff. You know, we talk about this all the time because if you provide that information to customers, you cut their costs. What do they do with that savings? Buy more stuff so they can innovate more. So that's the customer version of the flywheel, right? So yeah, so I wanted to talk to you a little bit more about Splunk Cloud. So you announced this at dot com. And so Splunk Cloud is running on AWS, correct? So talk a little bit about that offering. And specifically how AWS makes it possible. Yeah, I mean, we've actually had an offering in the cloud for a while. We have SplunkStorm, which is still available as a free service for developers. And a lot of the conversations that we've had with customers over the years, we are not unique in the sense that as customers, a lot of our large enterprises were moving more and more workloads to the cloud. And these are some of the things that AWS is certainly enabling and driving. They were saying we would like to consume and use Splunk in the cloud as a service. We love Splunk, we want Splunk as is, the full range of capabilities, the full platform with all of the REST API endpoints, the application model, alerting all of the features that we've really tried to put in place to deliver value to our customers. They wanted to consume that as a cloud service. So that really is what Splunk Cloud is. It's Splunk in the cloud, Splunk Enterprise in the cloud. All right, so some of your competitors, so we had Sumo Logic on yesterday and Logly, another competitor, both cloud-based, SAS-based log management applications from the get-go have kind of called you guys out and said, you know, Splunk Cloud really isn't a true SAS offering. It's not a real cloud offering. Try to set the record straight here. What do you take from that criticism and how do you respond to that? I mean, I think we really focus on what are we hearing for our customers and how do we deliver value to our customers? And it is a cloud offering in the sense that it provides the flexibility, the time to value, the sort of moving of sort of the burden of operations from the customer on-prem to the vendor. All the things that the cloud delivers, we're delivering that with Splunk Cloud. It's running in AWS. We are obviously partners and customers of AWS in that regard. And it's been really successful and it's been, you know, it really has, it's not been, you know, it's been received extremely warmly from our customers and we're going to continue to make investments in it. Well, I mean, I think it's somewhat understandable that those competitors would go after Splunk on that. You know, that's their value proposition is that we do what Splunk does, but we do it as a service. Well, Splunk got all started, right? I mean, you guys kick and butt. Right. And in fact, I think Logly is a former Splunker started. Maybe it was similar logic, but yeah. Well, it's, you know, successful IPO, great valuation. So everybody's now saying, oh, we could do that. Right. When you're top of the heap, you're going to get some arrows. So what gives you confidence that you can maintain the lead? This is the question. Well, I think, you know, and it's the, I think it's the options across deployments and consumption, you know, how can I, you know, because really we're about delivering the value of operational intelligence. There's all this data. It's coming from a bunch of different sources. It's all over the place. How do I put it in one place and how do I get value out of it and how do different types of folks get value out of that? Whether it's IT or the development organization or the business or sales. And so, you know, different requirements are there in terms of how do I consume that? So that's why, for example, we have Splunk Cloud, which is a service. And another thing we announced yesterday is both Splunk and Hunk available and Amazon machine, machine images and Amis in the Amazon marketplace. So for some customers, for their requirements that it may make more sense for them to say, yes, we would like to run Splunk in the cloud, but we would like to run it in sort of a traditional IS model, right? So they would, you know, have access to the virtual machine and manage the virtual machine. And we want to enable that. Splunk Cloud is a different situation where it's sort of the more traditional SaaS offering where it's like, I just want to show up with a login and my data and everything's going to work. And Splunk, you're going to make everything work. And that's really, you know, that's our investment there. And then, of course, we obviously have a fair amount of customers who are still and will continue to say, it makes the most sense for us and our use cases in our business to run Splunk on-prem or in a virtual private cloud or, you know, all of those things. And we want to be able to enable that sort of range of choice with our customers. So the Hunk on an Amazon machine image, the former example that you gave, that's what, a bring your own license model? Yes, so the Hunk and the Splunk Amis are bring your own license and that's Hunk on top of EMR. So that's, you know, there's a big value out there for folks that, let's say, they've decided, hey, like, I'd like to spin up some Hadoop clusters, they decided that Amazon's the right partner for us, they have data in EMR and they're looking for ways that, you know, I want to derive intelligence from that data. I want to explore that data and they're familiar with sort of how what Splunk has done in the traditional machine and IT data area. And that's really what Hunk is, Splunk analytics on top of Hadoop. So Amazon prides itself on pricing transparency. Yes. But when you go look at the whole pricing models on Amazon site, the bring your own license, one of the things it says there is that, for BYOL, the ISV is required and kind of encouraged, I'm not sure the exact language, it's kind of fuzzy to list on the marketplace, the AWS marketplace. And in that instance, Amazon's going to take, you know, they're 20% big and the ISV gets 80%. So are you cool with that as an ISV? We are extremely happy with our partnership with Amazon. It's been great for us and our customers. So no problem listening on the marketplace, but that's something you would be fine with. We've been really happy with our partnership with Amazon. We're going to continue to drive and do more interesting things like the cloud trail. So that kind of requirement doesn't scare you off at all. In fact, it sounds like you embraced that. Yeah, I mean, it really was, it's a tremendous channel for us. The same way we talked earlier about, sort of why I wear a rest of the wicket, we wanted to bring our technology and our value to where the developers are, think about cloud customers, how do we bring the value that we deliver on operational intelligence to where they are? Well, a lot of it is, they're in AWS. So that's interesting. So not every company's going to feel that way, I would predict, because you guys, you're looking at it as an incremental opportunity. Wow, there's another distribution channel for people that maybe don't know about Splunk. Listed on the Amazon marketplace, and it's also the right delivery model and use case for a certain number of our customers, right? We want to make sure that we enable and we empower across all of those use cases, so. Sensible, I'm not sure, Jeff Kelly, if Oracle's going to have that same attitude, but we'll see. Well, the cloud model totally screws up Oracle's business model, right? Well, but then again, you see Oracle's a partner, so it's- Well, absolutely. They have no choice. Right, I think that they're being dragged, kicking and screaming. They want to sell you a big box, put it in your data, and then charge you a lot of maintenance over the years. So, what are you seeing, John, in terms of EMR? What are you seeing for the uptake, what are the customers telling you about elastic map reduce? I mean, I think it's been in terms of the overall, how do you reduce some of the complexity around anything with new technology, right? I think EMR, it's still the early days of Hadoop, it's still the early days of big data. It still feels a little bit like, you know, assembling PC kits back in the late 70s, and you know- Popular mechanics. So, it's another way of like, how do you, you know, I'd like to use a big data solution, but I don't know how to use a soldering iron, right? So, that's sort of the EMR has been a great way. How do I, you know, spin up some Hadoop clusters, get data in there, make sure that I'm able to collect and capture all of this sort of big data that's out there. And I think where we're trying to deliver value is, okay, great, the data's in there. What are you getting out of it? How do you know what's in there? How do you derive value out of that? And that's, you know, whether that's a hunk on top of EMR or a hunk on top of Hadoop sitting anywhere else, that's really sort of the, and that's a request that came from customers, that's where we're trying to play and that's where we're trying to help folks. So, related questions, just curious to get your take on what you're seeing in the Hadoop market overall. So, you've got the, you know, the more, I guess if you can call it traditional vendors, the on-premise Hadoop vendors, Cladera and Hortonworks and MapR. Now you've got AWS with their EMR offering. How are you seeing that shake out? Are you seeing any trends around your customers looking more to EMR, talking more about Hadoop in the cloud? Have you seen any shifts in kind of deployment models of the way customers are looking at Hadoop? Yeah, I mean, I don't know that I could say that I personally or that we've found a specific winner in terms of deployment models or consumption models around Hadoop. I think what we've found and really where we're trying to deliver some value is it sort of feels like it's the, there seems to be a bit of a universal POC that everyone's sort of doing. Like everyone's just kind of kicking the tires, figuring out, all right, great, I had this in there. I read, you know, I've read that Big Data's a big deal and I have, I spun out some Hadoop clusters. Oh, my boss read Big Data's a big deal. I saw it on a blog post somewhere. But now, okay, here's the bits. What do I do with this? What's the value? How do I, how is this worth my investment? How is this worth my investment in terms of people and time? And, you know, you think about just the classic kind of technology adoption model is how is the, you know, the person that I'm going to hire tomorrow, whose job will largely be driven by, you know, acting on data in very near real time and making decisions based on that data. How are they empowered by this technology? And that's where I think we're still a little bit of a chasm on where we're trying to play there. Because if it's still a pretty small community of data scientists who have the, you know, the soldering irons of knowing how to bolt together a bunch of Apache projects, you know, it's really not delivering value at scale to the industry. And that's- That's the chauffeur analogy, right? You know, I'm talking about it. In the early days of the automobile, the big concern was, there aren't going to be enough chauffeurs. Yes. Yeah. There aren't going to be enough data scientists. Yeah. Well, how do you solve that problem? You either, one, create more data scientists, or two, you make the tools easier to use so you don't need a data scientist. Yes. And realistically, you do both. Yeah, and I think, and that's where we're actually, you know, looking at that exact issue in a couple of different ways. And I think, and I'll have a little bit of bias on the developer side, but if you think about, you know, Honk or even Splunk, individually, you know, what we did with Splunk 6 and some of the improvements to the UI and things like data models and pivot is, how do we enable sort of the folks outside of traditional IT to get data outside of Splunk? So again, so those folks are not chauffeurs for the data. So if I'm the business analyst, if I'm the person who classically would have been, you know, working in Excel, for example, and writing Excel macros, how do we enable those folks to do their job and get value out of that data? And so a lot of that is focused on the user experience and making that data user to use. On the developer side is, you know, again, back to the rest shirt, is how do we empower those developers with the skills they already have, right? And I think, you know, it's another great thing about the show, A, everyone knows what the shirt means, but you know, there's been such an explosion of sort of standards-based development, a little bit of a sort of a polyglot of different technologies, but restful is restful. It's sort of, it's a little bit table stakes, but it's important to deliver. Nobody's questioning the value of APIs anymore. No one's questioning the value of sort of standards-based web development. And so we wanted to make sure that, okay, let's say you have data sitting in HDFS, you have data sitting in Hadoop. If I'm, you know, you don't need it, you know, I'm sure there are scenarios for which you need to get a Stanford or Berkeley PhD to be a data scientist on top of it, but there are tens of millions of JavaScript developers out there who can build applications, who should be able to build applications powered by the data sitting, whether it's in HDFS or in Splunk, and that's really the core of the hard of the investments we're making in developers there. So there's already a huge community of developers out there where, you know, the code academies of the world are growing more and more tomorrow. It's sort of the new fluency, right? You have your eight-year-old, you got to teach them how to code. Well, how do we enable those folks with those skills? It's sort of the new, it's almost like the new information worker, right? Enabling those folks to get value out of that data without having to go through a graduate program. Well, in the early days of the PC, you would go take Lotus one, two, three training, right? I mean, I think the analogy applies, right? I mean, there's going to have to be some kind of training regimen, because it's, you know, pivot is great, awesome, but you just pick it up and use it, right? You got to have, you know, some training, but once you're trained, you could become, you know, do some serious damage. And then on the developer side, we really, you know, in Splunk 6, which we have this thing called the web platform, which is the layer that's backbone.js and jQuery, and these very sort of standards-based technologies driving it in Django, is we looked at it and say, we want building a Splunk application, whether that's an application on Splunk Enterprise, an application on Hunk, an application on Splunk Cloud, to look and feel like building any other web application, right? So if you have web development skills, if you have the sort of standards, model of you, paradigm of web development skills, great, you can build a big data app. And we feel like we're, you know, there aren't that many folks that are sort of saying that or delivering that right now, and that's really where we feel like we can deliver some value. Massive productivity gains, programmable infrastructure is obviously part of that, the whole DevOps movement, everybody wants to be on board, guys are at the center of it. And then in that way, there's sort of big data investments of, you know, 2011, 12, 13 of, I spun up a Hadoop cluster, doesn't become a kitchen appliance that gathers dust, because nobody knows how to use it. Absolutely, John Rooney, thanks very much for coming on theCUBE, it was a pleasure. Thank you so much, John. Great to have you. Thank you. All right, keep it right there, we'll be right back with our next guest. We're live, this is theCUBE from Reinvent 2013 from Las Vegas.