 Okay, we're back here live at the Fluent Conference. I'm John Furrier from SiliconANGLE. This is theCUBE, our flagship program. We go out to the events, extract the signal from the noise and talk to thought leaders, entrepreneurs, startups. Anyone who has the signal will try to share that with you. We're here for day two of the San Francisco for the Fluent Conference, put on by O'Reilly Media. But O'Reilly Media for great content, great data, all the streams go to SiliconANGLE.com. You'll see the streams there, as well as blog posts and ongoing coverage of the developer world. And I love going through these conferences for many reasons. You can see what developers are working on around the corner, get a peek around the corner and kind of connect the dots. And we've been doing that all week, but also want to talk to startups. And I think we heard from Splunk, which kind of evolved from a startup they're now a public company. We obviously Google's here. But we have a startup here, Concurrent Inc. Gary Nukumara, CEO of Concurrent. Just close their Series A funding with Rembrand Ventures in prominent VC in Sacray and Puneet at True Ventures. Congratulations and welcome to theCUBE. Thanks very much for having me, appreciate it. I love having startups on because startups have to be lean in me, but also work hard, build a good product market fit, and then go out and build that out, and then grow to your Series B and put the growth strategy together. That's your job as the CEO. You have a company that's been doing a lot of good product work in the seed funding. We've been following you guys on our side. We've been watching you guys. And now you have Series A and you're kind of putting it all together. You're playing in the big data space where developers are building apps. And it's hard to kind of describe. So first, Gary, talk about the update on Concurrent, the company and what happened with the funding, what's kind of going on with the company. And then talk about the marketplace because big data developers are really focused on analytics today, but it's fast evolving to integrating into legacy databases, kind of the service layers, kind of integration, a lot of interesting cloud and mobile integrations, all kinds of new things are happening. So talk about Concurrent and then talk about what you see in the market. Yeah, thanks again for having me. So Concurrent was founded in 2008 by Chris Wenzel, who's also the core contributor to the framework called Cascading. 75,000 downloads, very popular framework for building applications on top of Hadoop. It's bootstrapped until about 2010 and at the end of 2010, Chris went and got some seed funding from True Ventures and Rembrandt, which then brought us to present day. And recently we closed the series A to boost up the engineering effort and really focus on adding additional value to the users and the deployed applications. What is the product? What is the platform? Is it software development? It gets APIs? The front end, the open source product is a development framework. So it makes it really easy for, it's called Cascading and it makes it really easy for developers, people who know SQL or any data scientist to deploy an application on Hadoop in a very short amount of time. So we're taking the massive adoption that we have with the framework and we're commercializing some of the value, not the framework, but we're actually adding software value adds to it. So management, monitoring capabilities, be able to share applications, so on and so forth. So in the big data space, everyone talks about analytics, becoming the new programmer and that's simply more of the guys doing the BI kind of reports. But talk about the platform requirements for the enterprise relative to app developers because that is really ultimately where the puck is going. As they say, people are skating to where the puck is, Silicon Valley adage. So you guys are building a platform that is an advanced platform for big data applications. What is that market look like? What's the current state of app developers and what are you guys doing to build a product there? So big data, the business problem that big data created was very disruptive and Hadoop came in because it's a very economical way to store that data, scale it, access it, so on and so forth. Building applications on top of Hadoop was very difficult. That's what our founder, Chris Wenselsaw, very early on and so one of the things he did was he created the framework so that you would have a separate business logic layer from the data layer and he made it very logical so you can actually, it's written in Java, so any Java programmer can pick it up. It's logical, it's got a plumbing metaphor, so very logical from that perspective. The requirements for the enterprise is not to learn new skills to adopt Hadoop but to actually leverage existing skills and existing systems and existing investments that they already made in their infrastructure. So human resource investments and SQL or investments and then BI tools like SAS Institute or MicroStrategy, they want to leverage all of these things but also get the goodness of Hadoop and they also want to be able to actually very quickly build applications that represent their IP onto Hadoop. So one of the things obviously that we've seen here at the Fluent Conference and which by the way is put on by O'Reilly Media which you guys have a book coming out, we'll talk about in a second called Cascading written by Paco Nathan, one of your engineers and guys over there like Guru's over there at Concurrent. But one of the things that's coming out of this is the maturization of JavaScript which has been around for a while and it's not going anywhere but the API and the protocols of REST APIs really changed the game for developers. You're seeing a lot more server side, more database integration with JavaScript. How does your platform work into that construct? Because obviously everything's going to be API and as Dave Vellante, my co-host always says the future of the data center is going to be give me an API to the data center and that's kind of the way the mindset is moving right now into some of the tech elite circles. So how does your platform work with that API construct? So in the JavaScript world, we leverage JavaScript heavily on the back end as well as on the front end UI for a commercial monitoring product. And it's a, you know. But your platform has got APIs in it as well for developers? That's right. So there's a couple of different sets of APIs but to build applications there's a Java API, a SQL interface and a machine learning interface. You guys offer an SDK for developers? That's right. We do. The Cascading SDK comes with the framework plus a bunch of tools that you need to stand up on applications. So give me the process of how a developer would want to work with you guys. So give us the use case and the value proposition. I'm a developer. Why would I want to work with Concurrent? What would be some of the things I could do? So Concurrent, the framework actually is multiple things. It's the applications that you can build are kind of soup to nuts from ETL. So getting the data into, preparing the data and getting it into Hadoop and then extracting the data either via analytics or running models or so on and so forth. And the developer would interface multiple ways. If you're a Java developer, it's a simple Java API. You can just read the documentation and there's a set of APIs and you can program to it. And then if you understand SQL, if you want to actually hook Cascading up to an existing system or if you just want to write a Nancy-compliant SQL, then that's another way that you can unwrap onto Hadoop via Cascading. And lastly, if you're a data scientist and you want to run models on top of Cascading, you just export the model from your favorite tool, KMML, and drop it onto Cascading and it generates the necessary application to run your models on Hadoop. So basically it's an ease of use issue for you guys. You guys are proposing ease of use into that world. Yes, traditionally developing applications on Hadoop at the lower level APIs has been very, very difficult. And you'd have to learn new skills and you'd have to be highly skilled to actually go and do that. We're focused on making sure that it's easy for the mainstream enterprise to leverage Cascading and build applications and leverage their existing systems and investments that they made over the last 20 years. Let's talk about SQL because one of the things that was hot at Hadoop World then strata Hadoop World last year was SQL on Hadoop. It was like you got Impala, you got SQL, you got Adapt and a bunch of other companies coming out saying this is the big thing. Is that a big deal? I mean, it seems to me it's more of a kind of compatibility model for the existing enterprises. How big of a deal is the SQL piece? The SQL piece is very big. I mean, obviously there's, I don't know how. Install base. More than, it's 10X of Java. I mean, it's- Massive. It's massive. Data analysts can write SQL but can't write Java, right? So it's almost like English. It's a common language across the development world. It's been around for 40 years so it's- It's a must have. It's a must have for if Hadoop is going to cross over into the mainstream kind of skill set, SQL was absolutely imperative for them to do that as the first step. And we implemented it because it was very, it's very easy to, we actually took it a step further. We did an ANSI Compliance SQL, which is means you can actually hook it up to a JDBC driver and have a system talk to Hadoop. So we have a test suite internally that is something like 7,000 tests and two test joins and so on and so forth. And that's worked out great for us. I want to put a plug in for Instacray who's a VC at Rembrandt Ventures, your investor. Great, great technical VC. He still writes code. We're doing a segment later today on VCs who still write code and he's one of them. I see Mark Andreessen a few others. We'll talk about that but we always talk about the cloud being the mainframe and Paul Maritz talked about that in 2010 when he was at VMware mainframe of the future his software mainframe is back. It's the cloud. It's just in a different way. I had a CIO just tweet me, Donald Cox. I said mainframe computing 2013, question mark. Huge need, querying slash making demands for input for data. Don't forget processing needs of analytical tools. What does he mean by that? What's your take? If you read that tweet, okay he's saying mainframe computing question mark, okay. He's buying a little bit of the story but he's saying the huge need for querying and making demands for the information of the data and the processing needs of the tools. That's been kind of a complaint for Hadoop. Ben Batch, Cloud Air is trying to come out with Impala. He's right over there, what's your take? Given your experience in the industry and kind of where you are now on the cutting edge with Concurrent. That's that CIO saying I have needs. Yeah. What's your take on that tweet? Well let me give you an example of a use case where we've tied in the various use cases in the mainstream enterprise. So one of the big challenges of accessing this data, part of it is he's talking about its access, right? So okay so we invest in Hadoop, how do I get at it? And the problem is you gotta go hire the rocket scientists to go build the application to extract stuff out of it. That's why SQL is so important. Now what we've done is because we've implemented anti-compliant SQL you can take R or micro strategies, you can actually run a query through that tool. Now probably spent $10 million on micro strategies already. So you can hook a JDBC driver up to Hadoop via that, make your query, bring that into your tool, run, train your models, and then once you train that model you can actually then export that model back through Cascading and then run the model on Hadoop. You could do that in 24 hours. That's unheard of. Versus what would be the alternative? Go hire a Java developer and somebody who understands Pig and Hive months. Months, yeah. And testing is near impossible because Hadoop is generally a black box. And I think we had the Cube at Sapphire, SAP Sapphire show in Orlando a couple weeks ago and HANA's getting a lot of buzz because of the quote speed, now there's different use cases for HANA and we're researching that but they were talking about 26 days to 26 seconds to run a query. Now they're loading up a lot of stuff in RAM and again there's all kinds of arguments around that but again that's the kind of game changing notion. 26 days to 26 seconds. You're talking about potentially months to a day to run a complex data analysis. Right, so they're talking about latency. We're talking about time to market. Time to value. Time to go and look at your data, of extract value at your data. It still might be, we don't actually solve the latency problem. We solve the time to productivity, time to get your application deployed and go look at your data and start extracting value out of it. We solved that problem. But that's not a use case for latency. It's not like I need this report now. It's a pretty much known thing that they need to do. Yeah, right, cascading is more of a use case for we've invested a lot of money in Hadoop and we need to stand up a bunch of applications quickly. So gotta ask you, you guys are playing in a world that's very disruptive. Obviously the data warehousing business intelligence market has been kind of this fenced out, send the data out and months get some developers or run some pre-canned queries, bring the data back and try to work on it through other techniques. What's the current market for now? What's the use cases now where the disruption is on data warehousing and business intelligence? You're mentioning a little bit there. Can you share more data around or information that you can share around the current dynamics that are disrupting the data warehousing business intelligence market? The business intelligence side, there's many, many companies out there that are building that kind of capability in the cloud and many of them are actually using Hadoop and cascading to build that service in the cloud. So in the BI world, we're seeing a lot of that. On the kind of in-house data warehousing world, there are a lot of companies that have invested pretty heavily in the infrastructure that data warehousing infrastructure. They are looking for alternatives, cheaper, better alternatives to actually query their data. And you're talking orders of magnitude cheaper to store the data in Hadoop, but not orders of magnitude cheaper to actually get at the data and that's the problem that we're solving. So obviously for the folks that are watching that might not know SiliconANGLE, Wikibon, our research and our publishing side, we cover the hyperscale market and as INSIK knows, we used to be called web scale. Hyperscale is kind of the web companies. Now the enterprises are trying to do hyperscale and they're trying to get to the open source scale out without some infrastructure trends. On your website, I noticed you have a lot of logos there, Etsy, Twitter, Airbnb, are those customers of yours? Are those partners? Those are users. Those are users of your product, platform. They use cascading. All right, so those aren't like no names. They're big names. Etsy is probably one of the best big data examples out there. Pretty complex. Etsy's got 50 plus applications, cascading applications from ETL to funnel analysis to machine learning type applications that they have deployed. Twitter's got a dozen applications. They're like 10 times a day. Why are they working with you guys? What's the main aha moment? Because those names are an indicator to me that the enterprise follows those guys. I mean, we're seeing with open source, these guys have to build their own broad tools and use whatever's on the market to kind of do that. There's no off the shelf software for these guys. So they're a bellwether in terms of concerns. So why are they working with you guys? There's a couple of points here. One, from the enterprise perspective, I think the enterprises should take note of Twitter and Etsy and Airbnb and these companies that have used cascading for a number of reasons, but the one main one is that they were very early on on Hadoop and they went through all the pains of trying to operationalize and build applications that they needed to extract value out of the data that they were receiving and they evolved from there to saying, okay, well, we need to be able to separate the logic from the data and operationalize the application part of the extracting analytics or ETL, et cetera. And so the net of it is that they actually went through an evolution of figuring out how to operationalize their Hadoop infrastructure and they found cascading as a great way to do that because you can build it, test it and deploy it very easily. And like any open source project, we've been users of HBase and the trials of tribulation, it's been fantastic, it's also been frustrating. You want the promise, you get to see some immediate benefits and there's a huge maintenance issue, right? You got to wait for the open source to ratchet up the code base and you got also the demands of the business. Is that kind of the dynamic, not necessarily HBase, but is that kind of some of the dynamics that you're seeing that you guys are addressing is that, hey, these guys are adopters and they're using it and they just need better, robust, stable code or is it something different? They need that, but they need a more reliable way of building applications, that's the net of it. More reliable, logical way of building applications. There's an API, it's a set of APIs, it's not, they can read the code and they can understand what's going on. Okay, Gary, next slide, we're getting the time slot here. So the last question for you is, what's next for Concurrent? You guys have a great traction, and obviously Series A, Bootstrap company, I love this here, startups that are Bootstrap, this is fantastic, Testament to the Founders, you got two great investors with True Ventures and Rembrandt, your new CEO, you get fresh money, how much was the funding amount, can you talk about the amount, was it? Yeah, it was four million. Four million, so you got some good cash, not a lot, so you got to be lean, what's next for Concurrent? So we're focused on building the next phases, we're focused on building the value ads on deployments, so we've got thousands of deployments and tens of thousands of users of Cascading and they're very rapidly building applications, whether they're ETL or analytics. One of the big challenges is that they want to know what's going on with their application, did it start, did it stop, did it fail, what part of the process is it in, so on and so forth, so we're going to be able to extract that value and expose that to the user so that they have better visibility into the applications that they deploy. Okay, Gary Nukumaru with concurrentinc.com, Concurrent has Cascading, building an application platform for application development for big data in the enterprise and beyond, they have great clients, Twitter has great examples. Again, hot startup here at Fluent Conference, this is theCUBE, we'll be right back with our next guest after the short break and we'll talk to you shortly. I'm John Furrier from SiliconANGLE, we'll be right back.