 Live from Boston, Massachusetts, it's theCUBE at the HP Vertica Big Data Conference 2014. Brought to you by HP with your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We're here live in Boston, Massachusetts. This is theCUBE, our flagship program. We go out to the events and instruct the students in the noise. I'm John Furrier, the founder of SiliconANGLE. I'm joined by Coase Dave Vellante, co-founder of Wikibon.org. And our next guest is Andrew Rollins, chief software architect, co-founder of LocalLytics. Hot startup, not yet public, still private. Andrew, welcome to theCUBE. Thank you, thank you for having me. That's what we've got in there because, you know, you guys are in the later stages of finance now. So you see, you mature hundreds of people in the hot market, analytics, really on the cusp of really breaking out. We see this really early day still in big data. Exciting mobile. You know, if you think about the iPhone in 2007, we're still embryonic in terms of industry. The sea change just to mobile, just looking at web traffic, conversion to mobile. You have now multi-platform. Used to be set top box in the web. Now it's web mobile. Mobile first certainly changing the game. You guys are in the business of measuring analytics on the app, creating great engagement. Certainly a great value proposition as marketers look for omni-channel multi-user targeting, great stuff. So data's the heart of it. What's your vision? Big data, how to make all of this work seamlessly creating value for your customers? So yeah, that's a really good question. So we do more today than just the analytics portion. So we're focused on app analytics measuring how people use your application but also interacting with them. So that could mean anything from push notifications to in-app messaging, and maybe potentially other mediums as well. So it's not just measuring it but it's also doing something valuable with that data. And when you look at then the ecosystem as a whole you wanna do that not just on one device but you wanna do it across all devices, all touch points with your user. Whether that means the mobile app, the desktop, the website, maybe even on your point of sale system if you're let's say your retail company. So we're really looking into now, like you said in 2008 was when the iPhone app started launched, we're looking now into, well, forget just app by itself as a silo, forget just web by itself as a silo, how do we bring it all together? Unification is a huge thing right now and you brought up notifications and this is interesting because I think I wanna just drill down on this. What if I heard you correctly is most people have analytics that's basically monitoring basically what's happening and they report up on a dashboard. Yep, passive I guess, passive monitoring. Hey, we had visitors and show you some stats in real time but it's being aggregated in. What you guys are doing is different from what you said. In terms of you're interacting in a hand shaking way with users directly. So it's more passive active integration with the users. That's true? Yeah, so I could give you like in a use case. So the idea is really getting the whole feedback loop. So let's say you acquire users through some channel. Maybe you're advertising your mobile app on Facebook. So we measure the fact that that user was acquired by Facebook. Then we track their behavior and how they're using your app. What are they doing? And let's say you're a shopping application. You wanna say, okay, how are my users going through my purchase funnel? When are they adding an item? When are they going to check out screen, entering the shipping screen, the credit card screen, whatever. From there, you might want to identify groups of users that have made it so far into your funnel but let's say they've added something to a cart and they haven't come back to check it out. So what we can do is help you send messages, targeted messages to the users that have abandoned a cart. Try and pull them back into the application. So in that sense, we're trying to use the data here to really not just measure what's going on but do something about it. So we like to think of ourselves as not just an analytics company but a marketing platform. You're like an agent for like to aid, so I have my shopping cart. You don't know what I'm doing. I could have taken a break onto the bathroom, had lunch, gone for a walk. So like, that's what Pandora does when I'm sitting idle. Are you still listening? Kind of thing, you just ping the user, right? And interact with them. Okay, so now the value for the customer is convenience or kind of a feeling of interaction. On the analytics side for the marketer, what's the ROI for those guys? Do you bring data back in to look at statistics? Is there a holistic package? Is it more of an overlay network? We like to drill into the lifetime value of the customer. So your really overall, your objective is to increase, especially if you're a commerce app, increase what your LTV for that customer. What are you making off of that customer? How much are you selling them? For some other applications, it might be different. If you're a media app, you might be looking at engagement metrics because you're not selling directly to the customer. The customer is, your user is almost your customer than to an advertising company. So we like to drill into, let's say if you're in commerce, you're gonna really be looking at LTV. If you're in news or media, you're gonna be really looking at engagement metrics. How long are they in the application? What are they looking at? So when someone engages with you guys, they get involved right away. So is it mostly an e-commerce company? Is it an app developer? Is there a profile that you guys have for your customer base? Yeah, so we, these days play mostly in the larger applications. So many of the top 100 apps on iPhone and Android are on our platform. Our customers include companies like eBay, Microsoft, SoundCloud, a bunch of them. So not just even in the US, but globally. And we're really trying to work with app publishers who have long-term apps. I was reading the news yesterday as the Buzzfeed went public, not public, I'm sorry. Seems like a public offering. $50 million raising financing from Andres and Horowitz, a venture capital firm in Silicon Valley, and an $800 million valuation. Chris Dixon, one of the investors said, it's not about the category of them versus the New York Times, it's the fact that mobile now is, it's not about the category, it's about a whole nother generations of startups. So I got to ask you, what you're essentially doing is bringing in the internet of things concept where if everything's connected, the experience of a user is more than just an app or a web app or a website. There's a holistic view of this kind of full stack. What are you guys doing technically to make this kind of DevOps full stack startup work? I mean, is there a technology innovator that you guys have there? Is it just off the shelf guys coding? What are you guys doing? So what are we using for our stack in terms of how we're making this all come together? And do you guys believe in that thesis that the whole world is connected and that that's part of your vision and relevance? Yes, definitely. So for us, we really want to bring together one view of your customer that's cross-medium across all their devices. We like to think of apps, everything's becoming appified, the web too. It's not just mobile, in desktop as well. You look at what Microsoft did with Windows 8. So yeah, there is a kind of a uniform view. In terms of how we really deliver that, I mean, we're using kind of the latest and greatest tech out there. We're, Vertica is one of the things that we use, that's why I'm here at this conference. But yeah, and we're running in the cloud, we're running on AWS. We've gone through a whole bunch of different technologies in the six years. What's your main innovation as a company? What do you point to? Every company has their own little twist of uniqueness, Intel's, Moore's Law, secret sauce. What's your secret sauce in your company? What's that little edge? It's the real-time nature in being to interact with your customers on the fly, creating that feedback loop. I'd say our edge right now is in the mobile app and analytics and marketing space. We're one of the only closed-loop solutions out there where you can do both your analytics and all of your push notification, measure the results of that, see all your acquisition channels, all of that in one place and get a full-centric view. And like I said, not even just mobile really, but across any application web. So I'll use, I'll ask a different way. That's a great answer, by the way. I'm going to ask a follow-up on that. So Drop Can was bought by Google for half a billion dollars, the whole Nest thing. And essentially, we entered the founder at AWS Summit last year, great guy, awesome. He was struggling to get financing. It was a webcam connected to the cloud. No brainer, no one would fund him. Yeah, that's such a stupid idea. But what he did, he innovated on the storage aspect and provided an awesome SaaS product and everyone just scaled up and he used the time you pay as you want to store your video, who wouldn't want to store at least three days to check your video, and he made a ton of dough on one little innovation. So that was kind of his unique thing. Is there a technology thing that you're doing that's kind of your differentiator, that's your lever of success? Yeah, years ago, the typical kind of way of doing an analytics product in company from a tech perspective was, okay, I'm going to throw a bunch of data in Hadoop and then I'm going to kind of batch processes. So many of our companies, our competitors when we started were just doing batch processing. And they were four or five hours behind, sometimes even a day. We came out of the gate six years ago before anything was like real time and started to look at, okay, how do we do real time stream processing? How do we get you data within a minute of a user doing something and then be able to act on it right away? So that was kind of, that was what we led into the market with and that resonated really well. And we carry that forward now with all of the interaction stuff that we do, all of the messaging and notification and campaign management. And you do this predominantly with Vertica or using other databases? It's a lot of different stuff composed together. Vertica is great for pulling in the incoming data and munging over at real time and producing fast ad hoc results. So we have a question from our crowd chat. Go to crowdchat.net slash big HP, big data, 24, sorry, Dave, did I just interrupt you? Go ahead. So we have the crowd for Rollins, Andrew Rollins. Do you follow a DevOps practice inside your company and what does it look like when it comes to automation, deployment, delivery and so on? Oh, definitely. So given that we're entirely on AWS, we are very heavily into the DevOps mentality. There's a lot going on and I could talk for a really long time about that, but there's this movement now to make servers very ephemeral, very elastic so you can kind of up and down everything as you go. It's almost like every company is internally building their own Heroku and that's kind of the practice that we follow as well. We want to make it so that you can, all of our developers do not have to think about the infrastructure. They're just writing apps and they deploy them. So you have this whole kind of infrastructure layer that we build, almost an internal service for the company that is Heroku-like behavior. Using elastic beanstalk, Redis, things of that nature. All these kinds of things, yeah. A lot of different queuing layers, service discovery stuff, settling on a packaging mechanism like Docker. And I've got, interesting. And then I've got database services that I can invoke for whatever workload I need. So if I need fast ingestion, I can pull up my Vertica service. I presume you've got some in-memory stuff or maybe not. We do, yeah. And we actually make a lot of use of some of Amazon stuff like DynamoDB. And then you have this thing called the stickiness index. I presume you're not responsible for that. That's our marketing part of this. Our marketing thing, right? Yeah, yeah. But you're measuring it, right? Yeah, exactly, yeah. So you provide the data for the stickiness index. Exactly, right. That incorporates a lot of different metrics into it. And you sell this as a service, obviously, right? And you charge by month, based on how many users I have for my app? Yeah, see there, you can either price it based off users or the number of data points. Companies like to think in terms of users. And it's built for scale because you're doing Amazon, exactly. So a final question I have for you. I know we're coming on time. Since you're part of the DevOps, we're also big DevOps. We have big Amazon, our CrowdChat, our engagement container, fully instrumented with no big nodes, shop we are with the CrowdChat team. I want you to explain to the folks that are watching out there. There are a couple thousand people watching right now. Some are in the weeds and know our business, some don't. Explain the phenomenon of this DevOps revolution. The cloud revolution, why so much wealth is being created. Talk about it from a perspective of talent. There's born-in-the-cloud tech guys and there's non-born-in-the-cloud. That's my age, they're 40s, or maybe 30, 30 plus. If you're under 30, pretty much, you're coming out of school of computer science, you're native born-in-the-cloud integrated stacks, Amazon, you're not loading servers, loading patches, siloed. Talk about the culture and the differences between those two generations. What's the difference between a born-in-the-cloud developer and a non-born-in-the-cloud developer? This is something we debate all the time internally. So DevOps is an interesting term. And you have to realize that if you're near the field, DevOps is a loaded term. Because people sometimes hear ops and sometimes people hear dev and you don't want to get pigeonholed into one or the other. So you see a lot of different companies using entirely different terms. Like I think Google uses Site Reliability Engineer and some other companies use Infrastructure Engineer. That sounds a little bit more professional, a little bit more serious. DevOps, it's hard to get taken the wrong way because people look at you and they don't know. They think you're like a weird alien from out of space. Yeah. And sometimes they, the question goes through their head, well, are you a good developer or not? And are you just really an ops guy or not? So, I don't know. Or Renegade, unreliable, eating glass, hitting nails, running over other people. I don't know. I think if you're coming into the industry, do you want to be regarded as a really good engineer? Because the fundamental culture is managed through code, not through manual process, right? So you want to say that you both get what it means to manage infrastructure at scale, but you also know how to make reliable code that scales as well. Code is the key word. Infrastructure is code. Some people say infrastructure is code. That's really the DevOps movement. Infrastructure is code. You say that a lot. Infrastructure is code, programmable infrastructure. Yeah, abstracting away the complexity is getting stuff up and running. But at the same time, what he brings up is interesting, Dave. We've had people on that have been more from an ops background, instead of you call ops dev, because ops is more 5.9s oriented. Developers, oh, I just reboot both new code and the stacks take care. So there's more of a QA mindset, not so much one or the other, but more of it's your perspective. It's interesting those terms, infrastructure engineer and site reliability engineer. Doesn't that lose some of the development flavor? And isn't that where the sort of rock stars want to be? I think so. It comes in though in the engineer aspect. I think the word engineer, it's like there was a post on Hacker News recently that got voted up a lot, which said, don't call yourself a programmer. Call yourself an engineer. Basically, you should be calling yourself something else. And I think there's a little bit of an issue with DevOps where it can hurt people to call themselves DevOps people. I think you should look at it from an engineering reliability perspective. Well, I mean, you know, if you look at kind of the West Coast offense, where we're from, the days from the East Coast, your office is from the East Coast, a little bit more conservative. But like when the Facebook came in, it's like gunslinger mentality on the coding side. But even Mark Zuckerberg now admits his ethos has changed. It's move fast and make reliable infrastructure. They took out, it's no longer move fast, break stuff, because they're running at scale. That pie is not flying internally with their, but I agree. I think you're seeing classification on what I call IQ level engineering. You can be an engineer and not have to be pigeonholed and be bad, right? So if you're an ops guy and you're awesome at ops and you're engineering ops, that's not bad, but it's awesome, right? So if you're coding and you're an awesome engineer, you're writing great code, that scales as well as docking it, hopefully. So I agree, great comment. Give me the final word. What's your outlook for the cloud going forward? Real quick. I'll look for the cloud. I think containerization is a big thing. I think ephemeral immutable servers are a big thing. Being able to stand up infrastructure very quickly, things like mesos, I'm really excited about that. That's kind of where I see things going. We're beyond the instance-level stuff and into kind of a new wave of containerization. In a couple of weeks, we're going to have a crowd chat with mesos guys, Stackstorm, a guy who founded Nexenta, Evan Powell, those guys, Big Time, Alpha, Geeky, DevOps, Crew. So I'm going to email you for that if you don't, you'd be great, great addition to that chat. Very public, very independent. So, Andrew, thanks for spending the time. We went over again. You and CB kind of brought some energy last end of the day. Really appreciate it. This is theCUBE, we'll be right back with our next guest after this short break.