 Okay, we're back. This is John Furrier from SiliconANGLE.com, CloudANGLE.com and SiliconANGLE.tv at our new studio, the cube inside the cube right now at the Cloudera headquarters is Amr Awadallah, the co-founder of Cloudera. Take it away. Okay, so first I would like each one of you guys to briefly introduce yourself. Tell us your name, what you do at Cloudera. Maybe something fun about how you joined the company or about yourself in general. So Charles, you want to go first? Sure, hi. My name is Charles and I'm responsible for product management at Cloudera. I came here by way of SAP. A fun story about how I got here is Jeff Herrbacher found me on Quora. That's how I came to Cloudera. So that's par for the course for this company. So there you go guys. A strong tip for you to join Cloudera's to use Quora. Exactly. My name is Chad Metcalf. I'm an engineer at Cloudera and I'm responsible for build, test and release. I was recruited by another engineer here from a company that Mike Olson went to graduate school with the VP of Engineering. Kind of proving that it's just a really small world in the valley. My name is Eli Collins. I'm an engineer here. I also came via Jeff. He invited me to give a talk on some work I was doing at VM where I was for four and a half years before I came here. And just really hit off with the team. Really liked the management where the company was going and I've been here about a year since. So there you go. Another friend here is Jeff Herrbacher. One of the co-founders is a very good way to come into Cloudera. He hangs on Quora a lot as Charles said. So that's where you can find him. Okay. So the first question would go to Charles and Charles. I would like you to tell us how does Cloudera make money? You guys have this Apache Hadoop software which is open source and free. So how do you make money? Sure. So I mean a couple of things I'd say about that. So we have two products as a company. We have a distribution for Hadoop and then we also have a commercial offering. So even our distribution for Hadoop is a pretty significant step up over what any person or organization would get if they were to go to Apache Hadoop. So we make something that integrates a lot more components. It's kind of functionally richer. It's integrated, tested, easier to consume. So even our distribution is kind of a step up over what you would do if you were to sit down and load Apache. But on top of that, the kind of customers that we're serving, companies like Visa, eBay, Bank of America, having some bits is not really what they're looking for. They're looking to take this platform and apply it to all kinds of different use cases and business problems. And in order to operationalize that, they need a lot of other things. And we have a commercial offering. It's a combination of support and also proprietary software that we offer. And companies like this tend to find it pretty compelling to run Hadoop in production use cases. So you guys have some events coming up that we might hear from companies? We have Hadoop World. Hadoop World on October 12th in New York City. We've got five tracks running during the course of the day. I think total more like 30, 35 different presentations and panels ranging from kind of traditional enterprise to New World Web, to academic, to, you know, vendor stories, user stories. It should be a really good experience. So if you were to say briefly, what's the main value proposition of what Cloudera provides? Like why would I as a business buyer in a big bank like Visa or whatever? Why do I need Cloudera? Yeah, you know, the way we kind of describe it is that, you know, typically what happens is people read about Hadoop. They say, okay, I want to try it. Well, that's easy to do. You can go to downloads at cloudera.com. You can try it. And then they wind up using it for kind of a project, a single use case of some kind. And oh my God, it works. It's actually, you know, it's as flexible or as affordable or as scalable as I was told. And then they want to run it in production. And that's where our principal benefit is. We help companies run Hadoop in production. And that's a combination of proprietary software that we've developed as well as the team that supports this open source distribution that we've worked on for the past years. Thank you very much. So the next question is for Eli. And this question Eli is about what do Cloudera engineers do every day? What are they building? What problems are they trying to solve? So, I'm at a high level, what we're trying to do is build out what we think is a future for data storage and analysis. And as we've talked about before, Hadoop is a great starting point for that. In the platform team, we have Hadoop and then we have seven or eight other projects, many of which Cloudera created to kind of basically bring that future to existing companies. So people who see that core that's there and then they have a bunch of other problems that they need to solve in order to make that reality. So, you know, whether it's working on a distributed system to get data in and out of Hadoop in a reliable, high performance way or it's a single kind of compelling pane of glass that lets you look at a thousand cluster servers and know what's going on. Basically, any of the, we attack all the problems that are required for making that vision happen. So, if you can give me some examples of projects that Cloudera has contributed to. I mean, we know that MapReduce and the Hadoop distributed file system that Yahoo and Facebook, they did a lot of the contributions, most of the contributions actually in these projects. What are some projects that Yahoo, sorry, Cloudera has contributed to? Well, yeah, so we, like you mentioned, we collaborate with Yahoo and Facebook and others on core Hadoop. So I've, you know, done some work on the distributed file system. Doug and Tom and Todd and plenty of other people contribute to those core projects. But we also have some projects that we've created entirely ourselves. One of those is Flume, which is this, as I mentioned, we have a system that's distributed, available and reliable way to get a large amount of data in and out of the system. And some of our customers have, you know, five petabyte cluster, thousands of servers. You can't just run a simple crown job to load data in and get it out. That was a really interesting system. Some of our guys who here built and developed that system from scratch. We've got another interesting project called Yahoo, which is also something that Cloudera developed entirely on its own. It's short for Hadoop user interface, and that's an SDK for writing high level applications that run on Hadoop. So if you want to write an application to monitor a particular application, you can do that in, you know, an environment that gives you windowing and components that let you integrate with the rest of the Hadoop ecosystem, in terms of like looking at files and databases and that sort of stuff. So those are two very different kind of projects that we created in this ecosystem. And these are open source projects or proprietary? Yes, both of these are open source projects. So 100% of our platform is Apache licensed open source. And then as Charles mentioned, we have a bunch of proprietary products that we do on top of that. These two projects are open source. Yes. And you can find them on GitHub. You can find them on GitHub and github.com slash Cloudera. So the next question to you, Eli, still is three, two parts. First, if I'm an engineer, why would I want to come work here? And then second, what type of skills are you looking for right now? So I think the easiest way for me to answer that is why I wanted to work here. So I think in general, most people, most engineers I think are attracted by interesting problems and teams they want to work with. So this is very nascent technology. What we're building, we believe, is the future of data storage and analysis. And there's a ton of really interesting problems that have to be solved many years worth to make our dream reality. So I think A, there's a lot of interesting problems. B, it's a really strong engineering team. When I was coming, I was looking for a company that wasn't just trying to build like a widget for a website or whatnot. We were trying to look for somebody, I was looking for a company that could attack a really meaty problem that had an impact on real users. So if someone like Visa is using our software to do fraud detection or eBay is completely doing a bunch of new infrastructure on our work, it's really having an impact that's kind of larger than normal is another thing that I think is really attractive to engineers. In other words, working with stuff that makes a difference in the world. That's a very good point. So what skills are you looking for in the engineers that are coming to work here? So at a high level, I look for people who are passionate about software and systems and data. I think if you, you know, that's tend to be a pretty good kind of first pass at whether someone will be a good fit. We don't really hire for specific languages or frameworks. We look for kind of great engineers who can get stuff done. And a lot of them, you know, whether that's previous experience at other companies where they've done a lot of impressive work there or they're just coming out of school and have shown just a lot of promise during the interview, you know, we see that that ends up being materialized in a number of ways. We have, you know, all sorts of backgrounds on our team from experienced open source people, people like me who are at VMware for five years, you know, hacking and seeing assembly. So there's all sorts of background. So I think the common thread is that we're looking for kind of great engineers who have a really good taste in the type of problems they attack and work with customers and just want to write software that gets, you know, that solves real use and makes a difference and solves real people's problems. Very, very good. So maybe Joseph can give you a mic to Chad because you have a couple questions for him. So Chad, I heard that your group internally at Cloudera is called Kitchen. Yeah. Why is that? I wish it was because I would like to be a chef and one day work in the kitchen, but that's not the case. Actually, you are a good chef, right? I am a good chef and I have worked in kitchens and I now work in the kitchen. So when I first started here, I sat down with Jeff Homibacher and we talked a little bit about this team that we wanted to develop and we knew roughly what we wanted it to do. And so someone once asked a Google employee, why is Google so successful? And he said, it's because our kitchen is better. And what he meant by that was that the infrastructure and the environment in which they do their development is so much stronger and they've put so much time and investment into it. That's really what helps them succeed and kind of differentiate themselves. And so Jeff wanted something like that. And so we decided to go ahead and form this team called Kitchen. And Kitchen is really just a core set of engineers who are dedicated to really the hard problems of building, testing and releasing software at this scale. Like when you think about it, like CDH is, you know... Sorry, what is CDH? CDH is Cladera's distribution for Hadoop. So it's actually not just Hadoop, right? It's the entire ecosystem that sits around Hadoop. So we're talking almost a dozen projects. Each of them have various build systems. They all have unique test systems. They all have basically their own personalities. And now you want to build those. You want to be able to test them together as an integrated whole. And you want to release that. You know, that's a non-trivial problem. There's a lot of stuff there. There's a lot of cooking that goes into it. And so this is what the kitchen's really about, is, you know... Putting together, building, testing and releasing enterprise software. And realizing that it's a hard problem. It's more than just turning a crank, right? There's some hard problems and thought that has to go into that. So in Kitchen, do you guys just use custom software or do you guys build software as well? And what are examples of some software you're using right now? So it's a good question. I think a lot of companies run this risk of trying to be Google and building all their own infrastructure. And that's not the approach we take here. We do, because we are an open-source company, we do look for open-source tools that Phil needs. And if we can't find one, then we'll leverage existing frameworks to get where we need to go. And let me give you a couple of examples. So for testing, we use Hudson as kind of our continuous integration server. Right now we have a Hudson server running with 10 nodes, over 70 projects. It's continuously running a huge suite of tests for various projects, be it unit tests or integration tests or spinning up EC2 clusters to do packaging tests. Different versions of operating systems? Different versions of operating systems using various open-source tools to kind of look at the code quality, jcarter and find bugs and check style. And really, we leverage all of those. And then for some cases where there's not a system that really does this, there's no system that takes all the Apache projects and builds them magically for you. So we went ahead and rolled our own, but we started by leveraging some really important open-source tools like we use Bado, which is a Python binding for Amazon. We use that to kind of spin up Amazon clusters where we do all of our builds. We use S3 to kind of, one, hold our source artifacts and we actually use it to hold our binaries after everything was built. So everyone in KitchenRite software, it's not your typical release where it's just you push a button, you package up some software and you throw it on a website. Everyone on the KitchenRite software, everyone really believes and buys in on the open-source kind of community. We all work on various projects and contribute back to those. So basically we're just engineers on a slightly different non-release product. So then what type of... I was going to say, just to share another engineer's perspective on Kitchen is someone who largely consumes the work they do. It's important that you have a team devoted to quality, that putting out a quality product is really important to engineers and it's important that we devote people to that. And the other thing is that there's a lot of interesting problems in these kind of large-scale systems that existing ways of testing in building software just fall down. And you really need a fresh look and a really hardcore attack at those types of problems. So having a team of people who get to focus on that so that other people, for instance, if you're working on a change to the file system and someone's already written all this automation that will fire up a cluster on six nodes and let you start writing your software and you don't have to worry about those other things, it really means that other engineers can be more effective. So they both tack some interesting problems in terms of quality and scale and whatnot but they also really allow a lot of other people to kind of get their job done without having to get bogged down with a lot of things they don't want to get distracted by. So Chad, what type of skills are you looking for in the kitchen? I think the most important thing that we're looking for in kitchen are engineers. We want people who write software and who are interested in approaching kind of non-traditional engineering problems, right? So being able to look at a build system and see the very challenging sort of like hard problems there, that's what we want. If you look at a build system and think it's boring, mundane, kitchen's not going to be the place for you. But if you can see, we build for four different flavors of Ubuntu, Red Hat, 32 and 64 bit and you can see that this is actually a hard problem, right? Lots of dependencies. A lot of dependencies, a lot of different third-party software. How do you build these and how do you test them? We're also responsible for maintaining the clusters here at Cloudera. We have a development cluster. We have a CDH long live cluster. We have a CDH with security cluster. We have a dog food cluster where we run CDH to kind of analyze our own logs, our own data. So if you're interested in writing software that kind of helps maintain and monitor those that actually ends up being fed back into the various apps teams that do this kind of thing for Hadoop clusters, these are the kinds of skills we want. In the end, we're just an engineering team. Yeah, but beyond being an engineering team, you frequently would tell me that it takes a special type of person like yourself to be in the kitchen. What's the special extra thing that a kitchen engineer needs to have? It's not just technical talents. It's more like the way they approach how they do their job. Sorry if I'm putting you on the spot. No worries. That's what he told me. Kitchen people, so there's two different kinds of kitchen people. They're kind of more the operational sides of people who are going to work on the automation and spinning up the clusters. And for those kind of people, the more dev-opsies sort of attitudes are really what we're looking for. People that enjoy coding are kind of detailed focused. Don't want to keep doing manual tasks all the time is really important. And then the other side of it is definitely in my background is more the QA approach. And not people who look at QA as a black box, right? Like just check that they put an input in and get an output out. Like people who actually want to know what's in the box and kind of understand how they can potentially break the box. Those are really the two kind of people that we look for when we find people in kitchen. Cool. Thanks. So back to Charles if you can get in the mic. Sure. So first, Charles, tell us about what's the product development methodology slash process that you use here at Cloudera. Sure. So as I mentioned, we've got two products. And so at the moment, we have two teams. We've sort of changed that up based on what we're building at any given time. Our emphasis is on a natural process and with a lower case A, right? We don't have a lot of people with scrum manuals lying around. But I think the common sense part is that we want to build iteratively. We want to ship often and we want to get feedback as early as possible. And I think the other thing that I really believe in as a product manager is the shortest path from the customer's lips to the engineer's ears. And that plus the ability to ship regularly makes for a good product. So are you the only one who talks to customers or do you get the engineers to... No, that's exactly my point, right? So as many times as I can get out of that path, the happier I am, right? So I spend a lot of my time trying to recruit customers to talk directly to engineers rather than spend a lot of my time writing things down and, you know, I don't want to be a dog with a note tied to his neck. So you have a couple of open positions here at Cloudera right now. Can you tell us about them and what you're looking for? Sure. So we're hiring for a user experience designer and we're also hiring for a product manager. The user experience person, we're looking for someone who is familiar with interaction design and UI design. And I think this is really going to be our first UX hire and I think we'll play a pretty central role. So the way I've kind of described it is it's for somebody that maybe they've worked in some other organization as an individual contributor and they're looking to step up and be a leader and set the design look and feel for all of Cloudera products for the next years. With respect to the product manager, the way I describe that is, you know, we're looking for the best athlete. So I don't want to look for, you know, must have five years of storage system experience with data. I'm not trying to have somebody lateral in to do the exact same job they already did before. We're looking for someone who is focused on the market and the customer and the strategy for product first so that they're maybe as smart as our distinguished engineers who are not trying to out engineer the engineers because that would be a feudal project. And impossible, but that's impossible to do. So that's the kind of folks we're looking for. So you worked for a big enterprise software company, SAP, proprietary software as well. How different does it feel to work for an open source company like us? Yeah, I worked for SAP and I worked for BEA systems before that so two kind of brand name enterprise software companies. Honestly, the open source is a lot more fun. You know, I think in proprietary software, what happens is you hunker down, you build something, and then you spend a lot of time selling and marketing to get someone to actually try it, right? So if you look at a normal proprietary software company, to every dollar of revenue they bring in, they will spend 40 cents of that on sales and marketing three times as much as they spend on product. And you wait a long time, so you'll build something and it's like six months to go sell it and have it actually in use. With open source, you can get things out there faster. You can get feedback earlier. So the lag between what we do and when people get to experience the Clodera hotness is not very long. The other thing which is fun about it is that open source tends to show up more with platform technologies and platforms are all about ecosystems and communities. So from a strategic point of view, it's also a lot more fun, especially as a product guy. I think also for engineering, but we're playing chess here. We're not playing checkers, and there's a lot more levers to pull and there's a lot more interesting problems and puzzles to solve, and that always makes product people happy. Eli, you want to add anything about the open source nature? Yeah. You also worked at VMware, which was provided. So we have both people who work on open source and closed source software, which is kind of a role in both of our products. You know, it is interesting to work in an open source environment because you're collaborating with a lot of companies, not just internally with Clodera, but plenty of other companies, Yahoo, Facebook, Twitter, eBay, all these people that also work on these projects with us. One of the interesting things is that's allowed us to really grow much faster than we would have if we had not collaborated with them. So if you look at the types of customers, you know, we have, you know, within 12 months of launching the company, we had people running our software and hundreds of servers. We now have people running it on thousands of servers. It's really allowed us to kind of scale and tackle problems that companies our size normally don't get a chance to tackle. So I thought that's been one of the interesting experiences for me. One small thing just to add to that, I mean, just as a reference, this is not even a two-year-old company right now. And we have more than 50 paying customers on the backs of less than two salespeople. There's, you can't find a proprietary software startup that's had that kind of experience. I also had, and one of the, some engineers would say the open source nature allows them to share their poetry with the rest of the world, right? So as opposed to working for a proprietary company, only your peers in the company get to see your codes versus open source and everybody gets to see how beautiful your code is. So another question for you, Eli, Charles indicated earlier the Cloud Adder Enterprise and how we have the Cloud Adder Distribution for Hadoop, which is open source, but then we have Cloud Adder Enterprise which includes that, but also includes some proprietary tools that make it easier to manage and deploy Hadoop and monitor it, et cetera. How does that impact the engineering team? Is the engineering team now split between these kind of things or speak more about that? Yeah. So as Charles mentioned earlier, we had kind of organized our teams around products. So in general, we also just believe that small teams work. So we end up doing kind of each project ends up getting a small team that goes and attacks it. So we don't really, you know, the org chart is not actually something that really is actually sticky and sits there. It's how we happen to structure ourselves for whatever problems we face at the moment. So today, for example, there's one team called the Platform Team that produces all of the projects that go into our data analysis and storage platform. And then there's another team which is focused on, you know, the applications, kind of the system management and apps and ops that will go into making these clusters more usable to real users. And that's a separate team. And they were kind of a proprietary product. So we actually, I mean, we have an open floor plan and we all sit around each other. So you wouldn't actually be able to visually see necessarily that we're structured into those two teams, but that's actually how we're structured. It actually just made me think that I also remember to, you know, answer to your previous question about what's different about open source. One of the interesting things I've noticed in the Platform Team is we, since we actually work on the open source projects as well as our distribution of them, we actually do releases very quickly. So we release as the open source projects make releases. So we're involved both in the open source releases and the releases of our products. So, you know, from the time we start working on something to the time it's in, you know, we're getting feedback from the community and it's in users' hands, you know, in a matter of months, which you, like, you would not see in a traditional enterprise company. So that's another big difference with you with open source. So, Charles, I think this is the last question as for you. What do you think makes people excited to join and stay at Cloudera? Yeah, you know, I think Eli alluded to it at the beginning, so in part I'm just going to reiterate. I think that people want to work with other great-driven people. They want to work on big, challenging problems that are going to have real impact. And they want to be in a winning team. And I think Cloudera, you know, manages to offer all three of those things. Okay, thank you guys. Thanks, Amber.