 And from San Jose, in the heart of Silicon Valley, it's theCUBE, covering Big Data SV 2016. Now your host, John Purrier and George Gilbert. Okay, welcome back everyone. We are here live in Silicon Valley for theCUBE Silicon Angels flagship program. We go out to the events and extract the signal from the noise. We are here as part of Big Data Week, Big Data SV, Strata Hadoop, all going on right here in Silicon Valley. We're right across the street at the Convention Center where Strata Hadoop is king kicked off. I'm John Purrier. My name is George Gilbert, analyst at Wikibon. Our next guest is Kumar Sringani, who's the CEO of Blue Data. Welcome back to theCUBE. Great to see you. Thank you, John. Thank you. So, CUBE alumni, we've always had some great chats getting deep in the tech, but this is a technology road that's maturing. Yet there's still some new things coming in. Peter Burris and I were talking with George in the analyst segment around how the tech is being matured. There'll be some winners and some losers. New things are arriving. Now it's a spark of something we really haven't talked much about. So all this is kind of going on. So I want to get your take on what's going on and then compare that with how Blue Data is adjusting to all this because at the end of the day, we need full stack developers. We need faster compute and infrastructure in the cloud and I got to have real time data hitting the applications. Thank you. Blue Data is actually when we started, I mentioned that it's very important to how do the customers future proof their infrastructure. In fact, you can see over the last three years, we have journeyed from Hadoop to Spark and we even talking about Flink now. There's a lot of new technologies that are changing and how they're adding the value to what the customers are finding the insights on. Blue Data doing is very, very well. We are in the third year of our operation. We have dozens of customers all across the board, all the way from financial tech to pharmaceuticals to telcos and media companies. We have extremely good partnerships and a lot of these partnerships actually derived by the customers telling their vendors and say, you know, we'd love to work with Blue Data. A great example of that is we just announced a large partnership with Dell. We are certified. In fact, as we speak, they're introducing us to the lot of the customers because of the complexity that these technologies are bringing. In fact, customers are telling us, we don't even know which version of this stack works with, which version of this stack, and we have to spend a significant amount of energy. How's an entrepreneur out there in this market, how are you adjusting to some of the dynamics and how has that changed some of the things Blue Data has done? Be specific and you share some color around that. Yeah, I think it's very important that you focus on building the good product and finding the product market fight and iterate as quickly as you can while trying to spend as diligently as you can. I think if you can do those three things yourself, I think you will always find the signal, just like you say for the cube, you find the signal in the noise, you find the good customers. I think that end of the day, I am maniacally focused on the customer. If you solve the customer problem with the right amount of skill set, I think we will have a winning solution. We did not do any, even before all the changes that happened in February, our behavior is same. We still have the same set of people. We are focused on solving the customer problem. So tell us the sweet spot of the usage scenarios you're attacking and then let's explore from there. Yeah, very good question. I call this the customer, we see the customer journey. Typically what we are seeing is the departmental to DevOps, to production and to big data as a service on premises. Elaborate. So blue data was found on the single focus on trying to give a cloud like experience to the big data applications on premises. Because big data started as a bare metal centric and it is very rigid, very difficult. And if you recall my conversation with Dave, we call it create a cloud cluster in five most clicks. Hadoop was the very focus, but we conquered Hadoop. We now have all the distributions, Cloudera, Hardenworks, Pivotal, MapR on our platform. And we now have Spark, of course we've been having Spark more than a year and we are looking at all other applications. From a customer standpoint, whether he uses whichever application you use or whichever platform you use, you should be able to get up and running. That's the first step. And it will be typically starts as a departmental or a small cluster. Some big customers have a cluster sprawl and they would like to consolidate and do a multi-tenancy, reduce the total cost of the ownership, both on the capex and as well as on the apex. And then once you reach a point where you want to have a multi-tenancy and different people using your resources, you need a compute storage separation. You cannot have a multi-tenancy and big data as a service by tying the data locality to the compute any given time. Let me back you up a couple steps just to see if there's a way of measuring how much you can help complexity. Our research back last year showed us that customers were taking as long as two years to get from POC to production. And that the cluster sizes were like 10 to 12 nodes and that they needed four admins to run something like that. Tell us how you can change that equation. Our POCs are typically something like two weeks. All you need is a bare metal and the blue data software provides all the software that is including the distributions that are open source are prepackaged by us. You could be up and running in as little as a week and you don't need any expertise on what is the 1500 switches that you need on a particular cluster. Oh, so you can dial those in? There's a standard configuration. There's also under the hood advanced switch. You can open the advanced switch and you can set up. And you don't, this is the very focus for us. We just had a conference call with a $25, $30 billion company this morning with 12 people on the call. And this is they said, we have customers who called us and said, this is what we need. We don't want to have four Hadoop administrators slash and we just like to have a blue data. You can create a cluster. You can focus on what you need to do in terms of how you want to do your data analysis. I mentioned this in the last conversation of the day. I want to fundamentally change the data consumption model, right? Which is, I think you also pointed this to me when we were talking at this part. So for example, is it possible to put a number to, let's say, if your PSC is two weeks and let's say just to keep things, the server number comparable, let's say it's 10 to 12, how many admins would a customer be looking at if they're using you to deploy things? With respect to blue, I don't need more than one administration. I mean, you have a console. You can look at how many clusters are on that console. You can look at what is the usage of each cluster. You can look at how much CPU, how much memory, who is doing what, how long the cluster is up and running. You have an audit trial that is built into it. You don't need more than one administrator. Okay, wow. That's a big change. This is actually the point that we've been saying that we want to build that what we call if we brought the big easy button. In fact, in my last call with Jeff and Dave, I mentioned this and said we call it big easy and we're talking about New Orleans. So we want to bring that big easy data, big data easy button, which I think we brought to the marketplace. We have now done with it. We are now moving to all applications is also now packaged with the blue data. You've got an at scale application, you've got platform, you've got all the applications are pre-configured. Oh, so now, in other words, when you want to offer data as a service, it's, they can choose the tool, although they're responsible still for getting the data in in a way that's consumable. But everything then can be provisioned as though it were on Amazon's marketplace. You've got all the servers and then you've got the tools themselves, pick and choose. So you take bare metal servers, you pour blue data software and then you have this cloud like experience. This is why I said at the very beginning, think of us as a VMware for big data and we are now beyond even that. We are now the applications are configured. You got, we recently did a big webinar on Kafka, Cassandra provisioning on the top of the blue data. We provisioned Splunk, we provisioned AtScale, we provisioned Platformer and other distributions are important, but that game is over and whatever distribution you want, you can choose and you are done with it. Would this translate well to cloud environments? Extremely good question. So we are, so our focus has been on-premises because we felt that is where the data gravity is there. We are getting a lot of requests from the customers. There are some customers, they're collecting the data at both the places. We call it hybrid models and I think every distribution vendor who once upon a time said bare metal only have a novice strategy. So blue data is actually translates very well into the cloud. So we stay tuned for the next news for us. Hopefully in my next Q session we can talk about our next hybrid. Well Kumar, talk about the increase of new data types. I mean, we've seen all kinds of stories and we see big snapchats of the world. These big social platforms, engagement data. How's that gonna fit into your plans? Because as this new types of data comes in, the in-memory things spark is something that you invested heavily in. Place a role in that. What's your vision of becoming the easy button if you will, going beyond the easy button? You got the easy button. But as things get harder, how do you keep up with the pace of making things easy? So first we are not data translation experts. I want to be very clear. We partner with people like Informatica, Trifacta, and there are a lot of other good players that do it. We built a technological data tab. We discussed this in the last conference. Data tab is a, it gives the compute storage separation. You can connect to do any data sources. You can have your data on object store. You can have your data on Isilon. You can have your data on the HDFS. You can have your data on the local HDFS. In fact, we have a few customers who actually have an existing Isilon, and they have a compute store clusters, and they're connecting the data directly through the Isilon. This is the customers that are coming back to us. Now you have all of our compute clusters are stateless. You can shut them down, you can pause them, and you can reuse them. Going back to your question about, we are agnostic to how, where the data is coming from. You just tap the news source. I will admit that I am not an expert on, how do I get the data? But you're enabling some of the under the covers, under the hood. Kind of stuff we're saying, legacy stuff, and moving in from there. We, in fact, we are talking to another startup in the valley. They're doing this, the data translation, and they build these things, and it's a phenomenal opportunity for us to partner with them. So how would it work where we have some vendors who are doing really great work in turning Hadoop and Spark into a service in the cloud. Some are like maybe Amazon or more, lots of scripts trying to glue things together, and some have more friendly frameworks. But how would your offering work where they're really cloud only, we're gonna have some on-prem vendors offer cloud solutions or moving in that direction. How would you fit between those? So we think of us as the arms of players. So you can take a blue data software and you can create a blue data service experience on the top of any cloud provider you want. In fact, I have an example. We are working with outside United States, customer downloaded our piece of software, and he was very happy, he called us, and we actually have a POC running. But he happens to have hosting service with a provider. And that provider saw this service, and then he's now talking to us and see providing a service. So while blue data's business goal is not trying to build another big data as a service, competing with either with Amazon or others, we have the technology and the necessary things to focus on building it. But right now me as the co-founder and CEO of Blue Data, my focus is to give the customers best cloud-like experience on the premise so that he can serve his customers with the multi-tenancy, compute data separation, easy to use, and you don't have to wait three months to create a cluster and 42 administrators, as you pointed out, to manage that clusters. Okay, so this sounds like it's a big change in the OPEX component for big data. And the COPEX component as well. So, yeah. Oh, you mean because of the separation of compute and storage? The separation of compute as well as the underlying utilization of the CPU and the memory. So one industry survey points out in the bare metal, you're running as light as about 20% of your CPU. This is why we're working very closely with Intel. Intel is a big investor in our, in Blue Data. We work with Intel team. Doug Fisher, who is a senior vice president from Intel, he's on the board of Blue Data. We work very closely with Cloudera, Hardwareworks, and MapR. Our goal is to, if you actually tune Blue Data platform very close to the hardware, no matter which distribution runs, you can get that advantage of it. So you don't have to put the tune up in the distributions. You put the tune ups into the Blue Data platform. It's interesting that you talk about working so closely with Intel because in about the 2000, early 2000s, maybe 2005, Intel started putting instructions in the chip to help VMware. Yes, exactly. And it was kind of confusing because it basically meant you're going to use fewer chips because your utilization can improve. But from what we hear from Intel, they expect Hadoop servers in number, as part of clusters, to far outnumber database servers. So you're the new VMware for them. That's why I said at the beginning when people asked me, what is Blue Data? Even though I have a lot of respect for VMware and I'm from VMware, think of us like a VMware. So there's a bunch of things in what you said I like to elaborate on. Number one is all the new workloads are going to be around Big Data. And I'm using the word Big Data instead of Hadoop, whether it be Spark, or Plunk, or something else. Second thing is, if you can actually put all these optimizations into Blue Data platform, then you don't need to worry about putting into the distribution. So everybody benefits it. That's what Intel would like to see is Intel would like to see Big Data being used very effectively so that their chips and their things will go. Kumar, great to see you. I want to get your take to end the segment here on the show here, Hadoop World, Strata Hadoop, Big Data SV, your thoughts on the industry where it's at right now, you know, Progress Bar. Where are we on this era? I mean, we've got a lot more work to do, but still, it's value being created. It's a big theme here. What's your thoughts? No, I think it's an exciting time to be here, despite all what we have seen in the February timeframe. I think that there is a lot of enterprises and even the small companies are figuring out the value out of the data that is being generated. I think that between the IoT and the drones and the new data types that John, you pointed out, I'm actually very excited to be in this part of the journey and adding to this journey myself through Blue Data and giving the people the opportunity to make it easy for them to what I would call data consumption models. Great to see you. You're a great entrepreneur, great company, great family. Your son's also an entrepreneur out there doing some work at the AMP lab. Great to meet him at some of the events over the holidays. Great to see you. Thank you. And folks, Cube Madness Final Four has been released. Tanme, Bakshi, Bill Schmarzo, Sanjay Poonan, Colin Mahoney made it to the Cube Madness Final Four. Kumar, next year, we've got to get you up there. Get the voting machine going. I will try. Cube Madness, go to SiliconANGLE.tv, check the status of our March Madness, favorite guest voted by you, the community. We are here live at Strata Hadoop, Hadoop World, Big Data SV, this is the Cube. We'll be right back with more after this short break.