 Live from Orlando, Florida. It's theCUBE. Covering.conf18, brought to you by Splunk. Welcome back to Splunk's Conf 18. You're watching theCUBE, the leader in live tech coverage. I'm Dave Vellante with my co-host, Stu Miniman. We're here in Orlando, day one of two days of wall-to-wall coverage. This is our seventh year doing Splunk.conf. Stu, amazing show. A lot of action, partnership is growing, ecosystem is growing, and we're going to talk. One ecosystem partner, Gemini Data. JR Murray is here, he's the Vice President of Technical Services. Welcome to theCUBE. Thanks for coming on. Happy to be here. Yeah, so when we first started, the Splunk ecosystem was really tiny, and it's just sort of growing and growing, and now it's exploding. But tell us about Gemini Data. What are you guys all about? What's your role? Sure, so my role is VP of Technical Services. I manage our sales engineers and professional services consultants, as well as our managed services practice based in the United States. So what I do is I go through and help make sure all of these operations go pretty smoothly. And in terms of the company and what we do, we've got a couple different things that we work on. Primarily our focus is around big data platforms and making them easier to deploy and manage. We offer a hardware appliance as part of that package, and we also have an investigate software platform that we feed data into, and it helps analyst jobs be a little bit more easier and quicker to do investigations. And you guys started the company just three and a half, four years ago, is that right? That's right. Back when big data was, it kind of still is a mess. Doug even said that in his conversations today. He said, we live in a world filled with change. The messiest landscape is the data. That's right. And the bigger, the faster, the more complex the data, the messier it is. So you guys kind of started to solve a problem. That's right. Why did you start the company? What was the problem you were trying to solve? So really where we started is, we focused on there's a problem with deploying big data platforms. Customers have poor experience in terms of it's too complicated, there's a lot of very technical details you have to worry about. And if you're a little bit lower on the maturity curve of technology solution implementation, you might need some help along the way, or if you are a little bit further along in the technical maturity curve, you may actually need some help in getting something that's more turnkey in order to alleviate a lot of the challenges that go along with IT bureaucracy. You've got maybe something that you need that's purpose built because you've got something that's very central to your security strategy. You need to make sure that it's up and running and reliable and dependable. So that's where we come in. We have a platform that we allow you to implement it's turnkey solution, multiple systems, get your Splunk deployment up and running. And will you do that on your website, looking at support various technologies? I see Splunk on their FireEye, Cloud Era, ServiceNow, Amazon, Azure. So those are sort of systems, RSA. I mean, they've got a lot of products and in a lot of cases it's Cloud or they've got a platform like Splunk. Will you actually do like bottoms up stuff with Hadoop and Pig and Hive, or are you really focused on sort of that higher level, helping customers integrate those platforms that they've brought in? Kind of helping to be a platform of platforms, if you will, is it the form of the ladder? So that's the idea, right? We come in and we go through and we say, what are your actual goals here? Do you just want to go through and install Splunk or do you actually have a big data strategy that we can help you execute on? So it's kind of a cohesive, holistic approach in terms of what you need to deploy and how we help you get there. So if you need to deploy Splunk, we help you install Splunk. If you want to do Splunk and have a Hadoop data roll, for example, you can have Hadoop just alongside your Splunk all on the same platform. You can go through and manage that sensorally, make it a little bit easier to manage via policy, push out jobs centrally, all the automation orchestration is there and the underpinnings for all those solutions. JR, who are you typically selling to? One of the things we look at, data is pervasive in the company, in companies, but who owns it? I've talked to a number of people at this company that like, well, I've got Splunk and everybody comes and asks me questions now. So where do you fit in in the organization? So we've got a few different things going on. So in terms of who we sell to and where we focus, it's kind of across the board, right? We've got very large enterprises who are pushing tens of terabytes into the deployment and we help them out with getting a solution that's going to be something that's a little bit more manageable. You've got a limited staff, the knowledge of Splunk is hard to actually cultivate and then actually keep and retain folks that know Splunk. They are generally very well paid. So it's easy for them to find opportunities elsewhere. You've invested a lot in these people. Your success is very critical and they're a critical part of it and it's important to keep those people around. So we've got a managed service to help with customers like that. We call it GeminiCare. We come in and we are actually able to have an automated monitoring and break fix type of resolution service that factors into those types of deployments and as part of that we go through and offer some services, touch points throughout the month to make sure they're getting what they need from a value standpoint. I mean it's one thing to have the platform, the deployment, the data, but in fact if you're not getting any value out of that what good is it? So if you don't have the talent and the skills you're able to go through and use us to implement some of those use cases and things like that. Yeah, one of the other things that's changed a lot in the last three, four years is on-premises of course is where a lot of customers are and where a lot of data is but public cloud, you partner with the Azure's and Amazon's in the world even if you start talking about edge, that diversity of where my data lives. How's that playing into your solution? So it's funny you mentioned that. We came to market, we led with an appliance-based solution. We said customers are having problems either getting hardware, common thing is you want to put a box in or 10, 20 boxes but you've got the storage team saying hey you need to hook up to our sand. We spent millions of dollars on this, we're going to get some use out of it. Guess what Splunk? You're going to be our biggest consumer of all of our storage internally on this brand new sand we got. A lot of times it's not attractive to a lot of internal customers. You've got IOPS requirements, you've got all these other requirements. Folks don't understand that you've got hard requirements for CPUs and the bandwidth there. So if you're using virtual solutions which a lot of customers are forced into doing, you actually have a very difficult time getting reserved resources on those virtual hosts. You get a bare metal box in there, you get our platform on it, you have none of those issues. So in terms of where we pivoted from there, the industry is obviously going towards cloud. So what we're trying to do is actually, we have a solution in the market today. Customers are really interested in us helping them on that journey. So we've got plenty of customers who are on-premise today, they have a cloud strategy, they want to get out of the data center business and they need to get in the cloud. So what we're doing is we're helping them, we've got equipment in a co-located data center and what we're doing is migrating customers over to that infrastructure as more of a subscription basis. So it's the same platform, but now it's in the cloud. There are benefits to that. So I want to actually, let me follow up now. So the subscription basis, how does that work? So it used to be sort of an upfront perpetual license, and then here you go, and then we'll see you when there's another upgrade. And now, how's it working? I know 75% last quarter of Splunk's, I guess bookings or revenue, I'm not sure which one were subscription-based, ratable, and there was a big, long discussion about whatever, it was 606 and all the Wall Street guys trying to parse through it. What does it mean for the customer and what does that transition like? Is it, hey, good news, we're not going to go through these spike cycles, we're going to smooth things out for you, but what's that conversation like? We've got a lot of flexibility with customers. We've got the ability to do OPEX or CAPEX, we've got the ability to ship as an appliance, kind of an all-in-one solution. However, what we've really migrated to as what the market has demanded is, customer feedback is, hey, we can buy this box anywhere, and we're like, you know what, you're right. If you want to, go right ahead. Here's the software subscription. So now we have the option to sell the appliance and the software subscription together as one package that's also partially subscription, but what happens when you migrate that into the cloud is now you've got a cloud-based subscription infrastructure and that software license is sort of included in that. I want to ask you about use cases. We were talking a little bit before, go back before the term big data came to fruition. You kind of had the EDW was the so-called data, big data use case, and you had maybe a couple of analysts that do the decision support systems and can build a cube, and they were like the data gods. So big data comes in and you had use cases like a cheaper EDW, that was kind of a really popular one. Certainly fraud detection was the one, precision marketing, ad serving. Obviously Splunk in the security and IT operations space, although Splunk never really used the term big data, it was only sort of more recent. It lined up business analytics. So you're seeing all these sort of new uses for data, very complex as you pointed out. You guys started the company to sort of help squint through some of that complexity and actually build solutions. So the brief history of big data by Dave Vellante. So given all that, how has your customers' use of data changed over the last, since you guys have started and where do you see it going? So we originally started, originally we had some customers that came over into this new business venture, existing relationships and whatnot, they were using a different SIM platform. One of our primary objectives was to get them all into Splunk. And that's something that we were able to do successfully. So they were doing security analysis, log retention, and those were their primary goals and that's it. Maybe compliance, okay? So they're really focusing on that. Now today we're doing entirely different things. We're focusing on, as you mentioned, anti-fraud, huge opportunity in the space there with Splunk. You know, the tools in that space today are prohibitively expensive, very complex, and we come in with Splunk, we're able to take in data from all sorts of places and technologies, really no understanding of the data at that point required yet, and then we convert that into business value for the customer by means of services, because there's very little in the way of pre-canned use cases for that. And frankly, when it comes to the fraud space, a lot of customers, their requirements are all different. There aren't really many shops that are very much alike at all. So you've got to sort of manage around that. Now, that's one way, but we're also seeing folks who want to do executive reporting out of their Splunk data. You know, you're talking about being able to go through and do a year-over-year reporting. You know, how are we doing from a risk management standpoint? These are the things that you're starting to see trickle up to the C-suite in terms of what does that mean for us and the way we need to make these business decisions. So I want to say that. So it really started out kind of hardcore IT, certainly security use cases. What I'm hearing is Splunk is expanding into lines of business, actually using data in ways that, perhaps others were trying to do in the past, but not really succeeding. That's right. What is it about Splunk that allows you to do that? Have we heard a lot about 7.2 today? That's right. Performance improvements, you know, some efficiency, granularity of storage and compute. I'm sure the C-suite doesn't know or care about that. But being able to analyze more data is something that they probably would care about. Mobile is probably something that they care about. So what is it that Splunk's doing that maybe others, you know, aren't doing or can't do architecturally or technology? No, a couple of things stand out right off the top. So you've got the ability to scale. You've got horizontal distribution of data, which means you can spread that load across many, many nodes. We're able to go through and distribute that load and it makes things actually perform. So we get an acceptable user experience and that means everything to a customer, right? So that's one thing. The second thing is with Splunk, you've got Scheme at Read, you're able to pull in as much data as you want for as long as you want without having to understand that data. You can actually come back through later, you know, parse, interpret, report on, and get value out of that data historically without necessarily having to understand it upfront. That's, in my personal experience, been a huge impediment right up front to onboarding data with other, we'll call them legacy solutions, but there's still some in the market today that require and depend on that is knowing the data upfront. We can't pull in this data unless we know exactly what it's supposed to look like and can sanitize it, parse it into fields. So, Sue, I want to follow up if I may. So, you know, a lot of people in the big data world talk about no Scheme on Write or Scheme on Read, and what they do is they toss everything into a data lake. The big joke is the lake becomes a swamp. You know, they got to go and clean it up. Why is that not the case with Splunk? What's different about Splunk and that they're able to, I forget exactly how Doug said it, essentially structure the data when you need it. That's right. In the moment, I think he said. So, the difference with Splunk is that you're able to, you're able to foster and really pull together the community resources. You're more or less crowdsourcing how to parse all these data sources. You no longer have individuals at every given company with a very specific data source, say, Windows event logs that might be universal to many other applications and organizations needing to roll their own. So, you're able to socialize and share those things on a place like Splunkbase, and then suddenly everyone's able to really capitalize on their data. So, I see that as more like a force multiplier. You've got the entire community behind you helping you parse your data because they have the same data, and that's really what I think makes the difference. Whereas, the so-called data lake would be like the big data metaphor for a God box where only a few people know how to get to the data. Basically, yeah, that's right. And, you know, the amount of skill required. Okay, that's another big piece, is when you're in Splunk, everything's very, very well documented. So, if you need to write a search, it's, there are plenty of resources. You've got the Splunk community, you've also got all the documentation, you've got the quick reference sheets. It's not hard to get into. You know, it's hard to become an expert, but if you just need to do something very quickly, it's not that difficult. Well, if we look at where Splunk's going next, you talk a lot about the AI and the ML and one of the tensions you hear out there is, how much am I willing to let the system just take that action? So, I'm curious on your product line and working with Splunk, what you hear, you know, how real are the advances that we're getting with AI, ML, and deep learning and, you know, are users ready to embrace that yet? Yeah, so that's a technology that's truly made leave some balance even over the past five years, right? So, what we're seeing is customers are able to use machine learning to go through and do predictive analytics and to be able to have the machine sort of speculate as to, you know, and you could say predict, but it's really, I think, speculation more like, what a given, say, categorical value might be. Is it yes or no, maybe, for the answer to a question based on what those events say? Or is it, is there an outage coming up that potentially you can predict based on different values? And there are all sorts of applications for that and all sorts of platforms that are trying to do that. Now, what Splunk's done is sort of bring that to the masses with machine learning toolkit and made that a little bit easier to really digest for the common person. What they haven't done, at least until very recently, from what my understanding is that they're doing, is they're actually taking more of that functionality, making it more intuitive, helping customers understand the most common challenges, I'll say. So, you're really lowering the bar in terms of the amount of information or knowledge, rather, and skills to be able to leverage some of these more advanced algorithms and computing resources to go through and get the types of results you'd expect out of machine learning. Right. Well, JR, Mary, thanks so much for coming on theCUBE. Really appreciate your time. Great to meet you. Thank you. All right, keep it right there, everybody. Stu and I will be back with our next guest. You're watching theCUBE from SplunkConf 18 in Orlando. We'll be right back.