 Live from Dublin, Ireland, it's theCUBE. Covering Hadoop Summit Europe 2016. Brought to you by Hortonworks. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live in Dublin, Ireland. This is a European version of theCUBE. This is SiliconANGLE's flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier with my co-host Dave Vellante. My next guest is Rumerthi, the co-founder of Hortonworks. You are the CTO at one point. We now have a new CTO, but you're the co-founder. Arun, great to see you. Thank you, a pleasure. It's been a while, but it's fun to be back. Great to see you. You look great on stage, too. It was a great keynote. Really talking about the journey of Hadoop. It was nice to see the slides, a little throwback, 2006. Absolutely. It's been a while, isn't it? Through there, yeah. Tell us about, what's the world like for you these days, looking back like that? It's kind of amazing. One of the things I like about Europe is coming here. It gives you a sense of perspective on time. We talked about 10 years with Hadoop yesterday. I was at the Trinity College, and that one's been the first Provost they had was in 1592, so it gives you a sense of time. How small we are. Exactly. Hold it in my house. That was before the first pop, actually, which is actually. Yeah, so it's been an amazing journey. I couldn't have imagined how far it would come, but being here, we obviously want to go further. Hopefully, you guys and us will be talking in 10 years from now, and talking about the 20 years we did, not 10. Yeah, 10 years, congratulations. It's been a great run. It's been fun to watch you guys from the beginning, kind of spin out of Yahoo, whatever they call it, spin out, or take out, or whatever they call it. But you guys have done a lot of work, and what's interesting now is, under the new marketing and kind of the positioning, highlights kind of where you guys are going, this whole connected data platform. Absolutely. You know, as, we've always been focused as a, as a set of, as technologists and as a company, we've been focused on making sure we help our customers solve their business problems. Now, it's been pretty, it's becoming clearer and clearer by the day, especially over the last, I would say, 12 months. This notion of IoT is becoming, you know, important and real. It's been, we've been talking about IoT for a long time. You know, toasters and fridges for 10 or 20 years now. It's become real. I think there's a set of things which have happened. One is, it's getting cheaper and cheaper to move data around from the point of origin to its eventual destination place. You know, things like 3G and 4G definitely helps. Obviously, we play a part as a Hadoop community by making it really economically feasible to store and process data at scale, to curate data at scale. And last but not least, I think what we'll also see going forward is hardware itself. It'll get cheaper and cheaper to produce custom hardware, so you're going to stick hardware on everything, whether it's a Raspberry Pi, a $3 Raspberry Pi or whatever. So if you take these trends of cheap hardware or custom hardware, custom Asics, cheap communication channels and cheap storage and processing channels, it's the right, it's sort of the right point in time, if you will, from a technology curve to be able to solve this. And what we've also seen is, over the last couple of years, I'm sure, you know, there's the gardeners and everybody's talking about it, which is more data is getting generated outside the data center than within the data center at every enterprise today, whether it's social and mobile, you know, sensors, beacons, and so on. We, as a company, it was very important for us to, ensure we have a complete end-to-end story from the point of origin, which is where the Onyara acquisition came, and folks like Joe Witt, phenomenally smart guys, we love having them all at Hardenworks. It allows us to, you know, capture data from its point of origin into sort of classic analytics with Hadoop, and equally importantly, push the data back into the edge for, you know, better intelligence. And it's using the data for insights and actionable results. Absolutely. You know, one of these wiki-bonds talking about, and you mentioned Garda, but one of the things wiki-bonds talking about that I think is important is, taking the systems of intelligence concept to a whole other level, and they got a lot of traction with that, because the practitioners out there are looking at data, and Ragu from Microsoft talked about it, data is everything. You know, I wrote a blog post in 2007, saying data is the new development kit, kind of people like, what do you mean? It's now pretty clear that data is now part of the development process with DevOps and Cloud. Absolutely. Data is an ingredient for that. Absolutely, and you know, I think you put it well. So the way we see this is that every application that you're going to write in an enterprise is a data app. I mean, if you're not able to manage data, if you're not able to use data to derive actionable intelligence out of that, you're doing something less than ideal. So we feel like, again, if you go back in time, data is probably the most tangible asset you have in the enterprise. I mean, if you had a piece of software running in 1960, it's worthless right now, but if you had, let's say, an IBM stock code from 1960, it's definitely useful in some context or other. That's why data is the most tangible asset you have in the enterprise. The context is changing so rapidly that data's got to be fluid. Absolutely, and you've got to have lots of it, not just down sample data. Having data in its raw form is really important and we're starting to see that more and more of the industry sort of appreciated and actually take advantage. So since we're going down memory lane, I want to go back to the original founding days of Hortonworks. You guys used to have your, the original founders would have a statement that 50% of the world's data would be stored in Hadoop by whenever, 2015. Now, whether or not that happened, it's kind of irrelevant. The point was that was a sort of a unifying vision. And you're talking about, you know, the amount of data essentially that you can now economically store. Yeah, move and store, yeah. So have you sort of achieved the objectives of the intent of that statement? I think so, I think the biggest intent of that statement was to help people understand what we're talking about here, which is that data is the tangible asset they have and managing data and managing data is a skill that every enterprise needs to learn. And obviously the context has changed from the last 20 years or so. People who probably capture less than a person of data. However, if you look at what's happening in the world, probably more data has gotten generated in the last two years than his entire history of mankind, right? So I think that challenge is getting ever more larger and ever more important to solve, which is fundamentally being our thesis. And so is that your true North? Or is that just sort of plumbing and the true North is more sort of deeper business integration? Can you talk about sort of that current vision? Well, I think both are part of the true North. Like I said, people, you know, you cannot build software and applications in isolation, right? Which is why I think it brought up a great point which is having complete integration, whether it's security, whether it's, you know, technologies like the cloud, whether it's data governance, management and so on. Having that integration along with all the technology, along with all the investments you made in the enterprise, whether it's with Oracle or IBM or any of these guys, having that fully integrated version with in a secure fashion is super important. And that's how that helps us achieve the ultimate goal of managing data well and taking advantage of it. Arun, Dave first said when we did theCUBE when you guys came out and Claudio had a little bit of lead on you guys. Obviously that lead's changed completely around. Oh, Horton was going to be the red hat for Hadoop. I was kind of like, what people were talking about. And Dave would say, can there be a red hat for Hadoop? And that was kind of what people could put their arms around the open source paradigm. A lot of a lot has changed. Certainly you can't really say red hat was the Linux at a different time than Hadoop. We all know that, but brings up the point of kind of what you guys have turned into, which is essentially integration opportunity around open source. And this is something that we're teasing out on the Wikibon and SiliconANGLE team is we're identifying that most large enterprises want integration beyond some of the POCs or shadow IT land and expand adoption. And what does that mean for them? Because they want to integrate. And how do you guys help be a positive impact for that? So I think in our role now as one of the key, so you talk to any enterprise today that Horton works, probably it's the same for any of our competitors too. There's more data, or at least there's as much data in Hadoop as in any other system in the enterprise today, which means given the fact that everybody understands that data is important and it's tangible, we are a trusted partner to our enterprise customers, where we go in and help them integrate with not just their internal processes, but also with other vendors they may have, like a Pivotal, you saw the announcements today, Pivotal or an IBM or a Microsoft, we feel like playing that aggregated role of what through data is an important role, or an important sort of function we bring to the enterprise and we help them manage their data across all their systems, SAP, IBM, what have you. So I got to ask you the question around the startups who want to come in open source is really an organic environment. You guys run a great show here, it's really about the community, it's not a big commercial event that they have. I mean we're meant to keep it that way. The other shows do, but the integration's hard. So it's hard for a startup or even a company that's transforming to the cloud. Because of the scale involved, we heard Microsoft kind of tease it out by saying, hey we want to create these integration touchpoints. How has the bar changed for companies to win an enterprise customer with open source, with big data applications? What's the current things that you see as from a customer standpoint, what's the bar to the hurdle to jump over? So I think brought up a great point about Red Hat. I think we certainly, as a company, we have to be thankful for Red Hat because they showed that there was a business model around open source. And more importantly, we didn't have to sort of break new ground in terms of helping enterprise understand that there's a future with open source. A first class citizen, if you will. Exactly. Now we're now in sort of the privileged position of being able to sort of maybe help other sort of younger startups show that there's a future in this business. In terms of the integration thing, we see this too, which is we continue to want to make this out of the box integrated, whether it's with Microsoft technologies or Pivotal, because the bar right now is getting higher and higher, right? When I started on Hadoop back in 2006, it was just a tie project, right? So you could do whatever you want and people who liked it would download and run it. To fast forward that 10 years now, we occupy a pretty important part in the overall data stack at the enterprise, which means as Hadoop's become sort of the central, one of the more central pieces of your data architecture, the bar has gone up. Things like security and governance and metadata management and stuff. I never thought about 10 years ago because it wasn't important at that point. But right now, everything we build has to get secured, has to get integrated even within our own stack. Like when we started at Hardworks, we had probably the first version of HTTP, Hardworks Air Platform, the version one had about nine components. Today we have about 24, right? So every time we bring a new piece of technology in, we have to do exponentially larger amount of work to actually integrate it. And it's the same thing. I was just talking to one of our financial service customers, they're like, we love the fact that we can upgrade your software every six months because we trust you guys to do it. One of the concerns we have is if we upgrade your software, what happens to vendor X's software that is integrated with yours, why and so on? And it is a tough job. It's something that we continue to get better at, but it's something that we have to pay attention from an enterprise standpoint. So you talk about growing from nine to 24 components and a lot of that is sort of the original instantiation of Hadoop sort of hits limits. It's pushed to new use cases, so new components have to come out. Talk about how it's evolving. And there are some saying, oh, Hadoop is just a storage system. Okay, well, pretty good one actually. Lowered the cost, did some good things, but it's evolving. Well, I think, as with any technology, you either evolve or you die, right? I think Hadoop, for me, has become more of a brand in the ecosystem or a name for an ecosystem, but I sort of, one of my favorite examples is a ship of Tessias, right? Which says, if you take a ship and you rebuild it with every piece of wood, is it the same ship? I think that's the same with Hadoop right now. Is it the same ship? It really isn't, but it's still the same name, right? I think, fundamentally, every piece of that stack has changed. Hope, you know, sometimes, mostly for the better. Sometimes you do have challenges, but as a community and as a technology, as a set of technologies, it's something we are very aware of. If you don't change, you die, right? And if you want to be relevant two years from now, or even, you know, forget 10 years, but even two or three years from now, we'd have to continue to evolve. Well, Spark, I mean, memory and IoT are great examples of the evolution. Exactly. I mean, that's been a great direct impact. Absolutely. What's the impact of Spark for you guys? Where does it fit in the roadmap? I mean, other people kind of have positioned Spark a certain way. How do you guys do that? We love Spark. The fact that you can actually come in and do a more diverse set of workloads, whether it's SQL or streaming or whatever, using one API is really great. Now, having said that, we have to continue to push the bar. I mean, that's why we've been spending a lot of time internally on this notion of assemblies, which allows you to, you know, for me, good technology always fades into the background. It doesn't hit you in the face. Well, can you just back up for a second? Talk about assemblies. What is assemblies? And then go into... Absolutely. It's a good point. So for me, assembly is, if you look at what our customers are doing today, think of any application they're building, they have to put together a bunch of technologies. They have to put together Spark, and they have to have some NoSQL store, let's say it's HPS. They got to do some metadata management and security, which is NOx, Arrange, or any of these. And they want to maybe use a storm or whatever. It doesn't really matter. The point is we sell a bunch of these technologies, and the customer has to put it together himself or an ISV does it for us, right? Now, there's sort of fundamental limitations with in terms of skills and knowledge and sort of the effort it takes. And I think of it as sort of undifferentiated heavy lifting to put them over and over again, right? I mean, the fact that if you have to develop it on your laptop and then go to a testing environment and then re-prod and prod, you're doing sort of undifferentiated heavy lifting over and over again, the same stuff. And a new piece comes in, I got to test it again. Exactly, it's the revisions. Exactly. Now, wouldn't it be nicer if you go back to the Windows world? I mean, how many people wrote MFC applications? Very few, right? What people did was they downloaded an application, whether it was a notepad-like thing or word or whatever, and they downloaded and they ran, right? We feel like a similar thing needs to happen in the Hadoop ecosystem, which is instead of the customers putting these technologies together, we want a solution blueprint that you can download and run, right? And that's why it's got a specific configuration of components and how they're wired together. So, it's a runtime assembly concept, right? Exactly. So, you downloaded it, download it from GitHub, let's say. It's a simple spec. You pointed to your existing Hadoop cluster and Voila in sort of thinking in terms of- You are made a lot of the heavy lifting. Exactly. In sort of thinking in terms of technologies, you're thinking in terms of use cases. You're thinking in terms of, oh, I want a customer 360 application. I want a cybersecurity application. I want a farming application, if you will. So, you download, dot, and run, and sort of worrying about what's happening under the hood. So, in your slides today, you had the yarn.next, right? So, what is that? So, yarn.next is all about not.net. Not net, no. It's not next. It's not next. That would give me a trouble with that. That would be interesting. We have a lot to talk about then. So, for us, that's sort of the next phase of Hadoop. You can even call it Hadoop.next, right? Clearly, our marketing guys haven't gone into all of this yet. But- The power's never been. I'm sure. Now, the idea is, you set up this set of, you set up an API and a framework where you can now think in terms of a higher level solution blueprint, rather than in terms of individual technologies like Spark or Storm, right? So, for us, that's allowing those, you saw the spec I put up, right? I'm probably going to get crucified by putting code on a keynote. But that is a spec. You download the spec and you point it to an existing Hadoop cluster, you get an enterprise search application, right? We feel like that is really the next sort of phase for Hadoop, if you will. Like, the first phase of Hadoop was MapReduce, HDFS, then we got Yarn. Now, we've formed a focus on the app. So, take the first implementations and determine the recipes, basically. Exactly. So, the business impact is simplicity, speed, lower cost. Absolutely, solutions. Solutions, right? And we want to see hundreds of startups build those solutions and sell it. And so, what does it take for that to happen now? So, this is in what, you're just putting a spec out there? Yeah. So, in the community, this work has already started for the last, I would say, six to nine months. So, I would say, later this year or later half this year or maybe early next year, we get to a point where you can, I mean, if you go and download Hadoop code right now, you can actually play with it. But if you go from a product standpoint, we'd probably productize it in 2017. And this initiative that's in the community, vendors, customers together, what's the makeup of the- All of it. All of it, right? I mean, the promise for you as an enterprise is that you no longer have to worry about training everybody up in your enterprise on these skills. Sure, you'll definitely have some experts. But right now, to touch Hadoop itself, you need a bunch of skills, whether it's park or scholar, whatever it is. That's powerful. We want to make it so trivial that you can go to a vendor. They have a banking app or a 360 app or whatever. And you can download it. A lot of them will be open source. This allows people to the benefit to the startups and to our solution providers is they can come into these verticals with pre-packaged applications and compete and provide value. Absolutely. And make money. Exactly. We want them to make money. We want to provide the plumbing, but the way you put them together, I mean, a great example is the metron stuff they're doing, which is cybersecurity. We are building one of those apps too, right? Now, we're also building in a way where somebody, a startup would come, drive a specific machine intelligence app, a machine learning algorithm, and they can monetize it, right? We can do the plumbing that they can monetize it. So your role there in the context of building packaged apps is an accelerant. Is that right? Yeah, we want to build infrastructure apps so that there's a spec for you to build your applications. We will, in some cases, we will build the apps like Metron because we're forced to because there's nothing existing out there. But we want them to pump the market. Exactly. We are the catalysts. I mean, I think about Splunk. Splunk has some packaged apps that have been very, very successful, but that starts to bleed into your ecosystem. You don't want to do that too aggressively, obviously, right? We want everybody else to, we want, you know, a trissur or to bring an app, right? Like a banking app, they want to build it. Right now, it's hard. I mean, they have to come in, they have to plumb in, you know, edge base with high, you know. They're doing a lot of heavy lifting. Obviously, they're doing it a lot of heavy lifting. Well, it's hard because you need domain expertise. Absolutely. You need a lot of data science, expertise in the tech jobs. Yeah, and they want to basically minimize the lifting and focusing on the customer work. You guys take care of that for, okay, so we're going to ask you a personal question. So what are you doing now? You went public, company went public. You're the co-founder. You kind of, I mean, there's always pressure at the next level, gun to your head, public company, a lot of pressure. What are you working on now and what's happening within your daily life in Hortonworks? Well, more meetings, for sure. More customers. That's a good problem to have. You wrote a lot. Customer business. Probably, yeah. Certainly more than a couple years ago. Coding, doing reviews. All my coding happens when I'm on a flight. So I look forward to some of them because that's the only time I get to meet the customers. But I do get to spend a lot. I mean, we've got a fantastic team and a community. I get to spend a lot of time with them sort of continuing to drive the roadmap, but obviously I'm not executing in it. It's my team executing, but it's still, you know, a different level of satisfaction. You guys have oversight and advisory work. Exactly. All right, so here's the billion dollar question. So you're a successful entrepreneur. You've seen this from the beginning, still growing with the company. I'm sure you get a lot of questions from entrepreneurs. There's a ton of opportunity. You mentioned the crypto stuff and the security stuff. What are the good white spaces for entrepreneurs to sink their teeth in? What advice would you give for the folks out there who are on GitHub, contributing code, who get a band of brothers together or sisters, if you will, and just come together and say, hey, let's tackle something. Let's sink our teeth into some new, cool, but relevant areas. What advice would you give them? What navigation would you show? What path is there? I don't think there's one path, but what I've learned from my experience is details matter, pay attention to details, but also make sure that your customer doesn't have to pay attention to details. So you have to pay attention because it's important for you to get things right, but also make it sure that technology blends into the background, because I think the best piece of software are best piece of technology are those which blend into the background. And last but not least, I love the Steve Jobs Code, which says real artist ship, so ship's off there. Okay, so if you were going to do a startup right now. No way. What would you do? No, I prefer, if you were, what would you jump into? Of the ecosystem in the Hadoop ecosystem, what's compelling to jump in on a wave right now? I think, you know, data is still, I mean, I love the code from, I don't know, it's probably you guys, right? Data is a new wild kind of code. I think data is still interesting. There's also FinTech, which is interesting, and security. I was going to say. Yeah, security is not conducive. Leads me to my security question. So let's say, so Rob Bearden says to your CIO, listen, I don't know who your CIO is, I've not met him. But anyway, he says, listen, I want you to present to the board and communicate to them on security, like for us, our own internal security. So the CIO goes to you and says Arun, what do I tell him? What should the CIOs be telling their board of directors about security? Well, I think clearly security matters, because there are jobs in the line and legal stuff on it. But also as a CIO, it's important to understand that security is a means to an end. We constantly have this, lots of times we have constant dialogue with enterprises where security is almost used as, as something to beat people up with. A blunt instrument. Exactly. But I think, like I said, again, a great piece of technology, including security, they blend into the background, they let people achieve their end goal, but obviously make sure that they don't fall into a trap of doing something illegal or wrong. Yeah. What about the ecosystem? Communities win. Absolutely. We're seeing that. I think we're an example of that. It's a great post on Medium, the guy who started Secret, anonymous apps, identity matters, and that kind of brought out this whole, kind of realization that who you are and what you work on in public essentially matters. Certainly open source, it's been this way. Now more than ever, community, it will be the defining factor in winners and losers because the stacks developing, the solutions are coming online through runtime assemblies and new reference architectures. What is the community success factor? Has it changed at all? Evolved? Can you share any color into what's the current state of the community? I think, again, a lot of my experience has been colored by the ASF. I continue to be a big fan of that. Give you a very simple example. We started the Atlas project about a year and a half ago with a bunch of customers actually. You asked about how we do software. So we started the Atlas project a year and a half ago. Over the last six months, we've seen an amazing amount of involvement from IBM in Atlas. It's not something that happened at the corporate level. It's something that's happening at the grassroots level, which is amazing to see. Engineers within IBM. Exactly. Okay. That's great. We're doing this. Exactly. We're doing this over time. Their strategy evolved to the fact that they now are big supporters of Atlas and they continue to contribute lots of their IP and software and time. So again, it's just a small example of how, even though obviously IBM has their distribution, we have ours, we still can collaborate in the community the same way we continue to collaborate with the cloud data on Hadoop or with Pivotal or whoever it is. So I think I'm a big fan of the community model, especially the model that the ASF brings along. I think that's here to say. Okay. Your take on the European show here, obviously it's always been Europe was different in the States, different requirements, obviously different privacy rules are pretty obvious. But as the community gets global, you got Asia Pacific, Europe, South America, Latin America, everything's kind of all now one black world. Are there similarities now? Is it blurring? Is it still different? Can you compare? I think they're getting more and more similar. In fact, some of our customers here in Europe are actually driving on aspects like security and governance even harder. Like you brought up the fact that data from France cannot leave its borders. So we've had to build a bunch of technologies in Atlas and Ranger to do things like geo-based policies. That was not a concern in the States, but we brought Atlas here and Ranger here and immediately we were told you have to do this. And that was great because it collectively allowed us to make our software even better. So I love the fact that we get, there's obviously a lot of similarities going forward as the markets mature, but I love the fact that we're getting newer and newer interesting and more interesting requirements out of Europe too. So I'm a big fan. Final question. What are you most excited about right now in the ecosystem, the tech? You mentioned Akka's Rift, obviously we put it on Facebook before we came on. Obviously AI is great, cognitive is good. You're seeing kind of the coolness of computer science around some of the futuristic stuff that we all dreamed about. What are you excited about that stuff? What other, can you share what you're excited about right now? In the Hadoop space or more general? More general. Yeah, in general, yeah. So growing up, my dream, I mean I start programming on a stand. So my pro, and I start playing chess around the same age. So my growing up, my dream was to build a chess playing software. But then at some point I learned Go, right? And I never imagined that, you know, you could write software to beat a human at Go. And here we are, right? It was, I'm blown away with what, you know, Dennis Hassibis over here in London has done. I think that's massive. I'm a big fan. I think it's a moment to celebrate for all of us as engineers. So what do you value guys think about that in the context, you know, Watson had this big grand challenge, right? And then now Go takes it to a new level, right? I mean, so, I mean, what's the take in the valley on Watson? People are geeking out. I mean, I think it's really wonderful. I mean, is it competitive to what Google and Facebook are doing? Is it just old? Well, there's an arms-raised aspect to the marketing side of it, though. But I think there's very few, and the population is growing, computer scientists are around this area. So it's just more promotion. I mean, the challenge of Go is significantly higher than chess. Now, Jeopardy, different situation, right? Well, I mean, things have obviously, it's a testimony to how technology can use to improve, right? I mean, one of the big things that Watson did really well was natural language processing. Now I've got a $120 device at home called Amazon Echo, which is probably as good when it comes to natural language processing, right? So it's pretty, it's an amazing horizontal applications as you were saying, John. I mean, Dave, I think people, there's a whole new wave, at least I was talking with Cloud Foundry CEO about this. We're older in the computer science, a little bit older than Arun, but when we were growing up, computer science, we kind of knew everybody pretty much now. The kids are natively embracing computer science as oxygen, and there's a natural thirst for geeking out on writing code and or either being involved in some sort of building or something. So I think when you start to see this automation facilitation around extracting away complexities, people get jazzed about it. I mean, I think Silicon Valley is going nuts. I think Facebook is a clear example of the Oculus Rift. You got the new cameras, you got Google self-driving cars. You have all kinds of cool things going on, Amazon with Echo. So I think, Dave, I think it's just just the beginning of a renaissance of computer science that we're going to see explosion. Well, being able to sort of access that tooling, that sort of AI tooling, NLP, whatever, in everyday systems is huge. When you can get non-programmers becoming programmers, that'll be the secret of the action. And essentially it's almost like 4GL back in the day. Is that something that you see as a new term reality? Look at a couple of things. Even look at things like 3D printers. Pretty much every kid is going to grow up with one pretty soon, right? Not with things like AI. Who knows what's going to happen. I mean, the opportunities are sort of endless. The robotics clubs now are almost in every high school. You're seeing robotics exploding. Did you hear that story on NPR the other day about the kid, some college kid, who basically straightened his teeth with a 3D printer? Wow, really? He basically conducted ortho on himself. That's taking some good cash. I like to figure out how to hit my kids on the program. You know, a lot of people do it themselves. I got to call this guy. Believe me, I got four kids. I know exactly the racket called braces. And the reaction from the industry, the entrenched industry is, well, is there a good thing there wasn't bone disease in there, because that could have been, come on. He's just pretty amazing. Pretty amazing. Talk about disruption. Arun, great run. Congratulations on all your success. Great to have you on theCUBE, sharing your insight and experiences and just overall take. Appreciate it. And thanks for supporting us. Thanks again. Appreciate it. Definitely appreciate it. Good to see you. We are live here in Dublin. We'll be back with more live coverage here of Port and Works Hadoop Summit. Real great community. Not a big commercial show. Really getting down and dirty. A lot of education. And I bring the European flavor to it. So some great t-shirts day with, you know, the elephant with the Irish outfit on. Slauncha. I got to get, I got to get that. We'll be back with more after this short break.