 Live from New York, it's theCUBE. Covering theCUBE, New York City, 2018. Brought to you by SiliconANGLE Media and its ecosystem partners. I'm John Furrier with Peter Burris, our next two guests are Jim Franklin with Dell EMC Director of Product Management and a new gentleman, Andy, who is the Vice President of Product at Blue Data. Welcome to theCUBE, good to see you. Thanks, John. Thanks for coming on. I've been following Blue Data since the founding great company and the founders are great, great teams, so thanks for coming on and chairing it. What's going on, I appreciate it. It's a pleasure, thanks for the opportunity. So, Jim, talk about the Dell relationship with Blue Data. What are you guys doing? You have the Dell Ready solutions. How is that related to now? Because you've seen this industry with us over the years. Morph, it's really now about the setup days are over. It's about proof points. That's right. AI and machine learning are driving the signal which is saying, we need results. There's action on the developer side. There's action on the deployment. People want ROI, that's the main focus. That's right, that's right. And we've seen this journey happen from the Hadoop batch processing days and we're seeing that customer base mature and come along. So the reason why we partnered with Blue Data is you have to have those softwares, you have to have the containerization, you have to have the algorithms and things like that in order to make this real. So it's been a great partnership with Blue Data. It's dated back actually a little farther back than some may realize all the way to 2015, believe it or not, when we used to incorporate Blue Data with Iceland. So it's been actually a pretty positive partnership. Now, we've talked with you guys in the past. You guys were on the cutting edge. This is back when Docker containers were fashionable, but now containers have become so proliferated out there. It's not just Docker, it's containerization has been the wave. Now Kubernetes on top of it is really bringing in the orchestration. This is really making the, kind of the storage and the network so much more valuable with workloads, with respect to workloads and AI as part of that. How do you guys navigate those waters now? What's the Blue Data update? How are you guys taking advantage of that big wave? Yeah, I mean, I think a great observation. We embrace Docker containers even before actually Docker was even formed as a company by that time and Kubernetes was just getting launched. So we saw the value of Docker containers very early on in terms of being able to obviously provide the agility, elasticity, but also from a packaging of applications perspective. As we all know, it's a very dynamic environment. And today I think we're very happy to note that with Kubernetes being kind of a household name now, especially tech companies. So the way we're kind of navigating this is we have a turnkey product which has containerization. And then now we are taking our value proposition of big data and AI ML lifecycle management and bringing it to Kubernetes with an open source project that we launched called Cube Director under our blue kit as umbrella. So we're all about bringing stateful applications like Hadoop, AI ML to the community and to our customer base, which is some of the largest financial services and healthcare customers. So the container revolution has certainly gripped developers. And developers have always had a history of chasing after the next cool technology. And for good reason. Developers tend not to just chase after the shiny thing. They chase after the most productive thing. And they start using it and they start learning about it and they make themselves valuable and they build more valuable applications as a result. But there's this interesting meshing of creators, makers and the software world between the development community and the data science community. How are data scientists who you must be spending a fair amount of time with starting to adopt containers, what are they looking at? Are they even aware of this as you try to help these communities come together? Yeah, so we absolutely talk to the data scientists and they are the drivers of determining what applications they want to consume for the different use cases. But at the end of the day, the person who has to deliver these applications, data scientists care about time to value, getting the environment quickly all prepared so they can access the right data sets. So in many ways, most of our customers, many of them are unaware that there's actually containers under the hood. This is the data scientist. The data scientist. But the actual administrators and the system administrators who are making these tools available, making these tools available, are using containers as a way to accelerate the way they package the software, which has a whole bunch of dependent libraries and there's a lot of complexity out there. So they're simplifying all that and providing the environment as quickly as possible. And in so doing, making sure that whatever workloads are put together, can scale, can be combined differently and recombined differently based on the requirements of the data scientist. So the data scientist sees the tool. The tool is manifest in concert with some of these new container-related technologies and then the whole CICD process supports the data scientist better. The other thing to think about though is that this also allows freedom of choice. And we were discussing off camera before, these developers, they want to pick out what they want to pick out, what they want to work with. They don't want to have to be locked in. So with containers, you can also speed that deployment but give them freedom to choose the tools that make them best productive. That'll make them much happier and probably much more efficient. So there's a separation under the data science tools and the developer tools, but they end up all supporting the same basic objective of building. So how does the infrastructure play in this? Because the challenge of big data for the last five years, as John and I both know, is that a lot of people conflated the outcome of data science, the outcome of big data, with the process of standing up clusters and lining up a dupe. And if they failed in the infrastructure, they said it was a failure overall. So how are you making the infrastructure really simple and line up with this time-to-value story? Well, the reality is we all need food and water. IT still needs server and storage in order to work. But at the end of the day, the abstraction has to be there, right? Just like VMware in the early days, clouds, right, containers with blue data, right, is just another way to create a layer of abstraction. But this one is in the context of what the data scientist is trying to get done. And that's the key to why we partnered with the blue data and why we delivered big data as a server. So at that point, what's the update from Dell, AMC and Dell, particularly on analytics? Obviously you guys work with a lot of customers that have challenges. How are you positioned? How are you solving those problems? What are those problems? Because we know there's some AI rumors coming up. Big Dell event coming up. There's rumors there's a lot of AI involved. I'm speculating it's going to be probably a new kind of hardware device and software. What's the state of the analytics today? Yeah, so I think a lot of the customers we talked about, they were born in that batch processing, that Hadoop space we just talked about, right? I think they've largely got that right. They've largely got that figured out. But now we're seeing proliferation of AI tools, proliferation of sandbox environments, and then you're starting to see a little bit of silo behavior happening, right? So what we're trying to do is that IT shop is trying to dispatch those environments, dispatch it with some speed, right? With some agility. But not necessarily, they want to have it at the right economic model as well. So we're trying to strike it better balancing, hey, I've invested in all this infrastructure already, I need to modernize it, and then I also need to offer it up in a way that data scientists can consume it. Oh, by the way, we're starting to see them start to hire more and more of these data scientists. Well, you don't want your data scientist, this very expensive, intelligent resource, sitting there doing data mining, data cleansing, ETL offloads, we want them actually doing modeling and analytics, right? So we find that a lot of times that right now, as you're doing the operational change, the operational mindset, as you're starting to hire these very expensive people to do this very good work, at the core is still the data, but they need to get productive in the way that you hire them to be productive. So what is this ready solutions? Can you just explain what that is? Is it a program? Is it a hardware? Is it a solution? What is the ready solution? Well, generally speaking, John, right? What we do as a division is we look for value workloads, just generally speaking, not necessarily in batch processing or AI or applications, right? And we try and create an environment that solves that customer challenge. Typically they're very complex, right? SAP Oracle databases, AI, my goodness, right? Very difficult markets. Right, the tools are using hives, no sequel, all this stuff's going on. Cassandra, yeah, TensorFlow, you've got all, so we try and bid together a set of knowledge experts. That's the key for what, the intellectual property of our engineers and their deep knowledge expertise in a certain area. So for AI, we have a set of them back at the shop, right, they're in the lab and this is what they do and they're serving up these models, they're putting data through its paces. They're doing the work of a data scientist. They are data scientists. And so this is where blue data comes in. You guys are part of this abstraction layer in the ready solutions offering, is that how it works? Yeah, so we are the software that enables the self-service experience, the multi-tenancy, that the consumers of the ready solution would want in terms of being able to onboard multiple different groups of users, lines of business. So you could have a user that wants to run basic spark clusters, spark jobs, or you could have another user group that's using TensorFlow, accelerated by a special type of CPU or GPU, and so you can have them all on the same infrastructure. One of the things, Peter and I were talking about Dave Vellante, who was here yesterday with me, he's at another event right now, getting some content, but one of the things we observed, and this is, we saw this a while ago, so it's not new to us, but certainly we're seeing the impact at this event. The Hadoop world, which is now called Strata Data WorkCube NYC, is that we hear words like Kubernetes and multi-cloud and Istio for the first time at this event. This is the impact of the cloud. I mean, the cloud has essentially leveled the Hadoop world, certainly there's some Hadoop activity going on there. People have clusters that are standing up infrastructure for analytical infrastructure to do analytics, obviously AI drives that, but now you have the cloud being a power base changing that infrastructure, analytics, infrastructure. How has that impacted you guys, blue data? How are you guys impacted by the cloud? Tailwind for you guys, helpful, good? No, I think you described it well. It is a tailwind, right? This space is about the data, not where the data lives necessarily, but the robustness of the data, right? So whether that's in the cloud, whether that's on premise, whether that's on premise in your own private cloud, I think anywhere where there's data that can be gathered, modeled, and new insights being pulled out of, this is wonderful. So as we digitize assets, as we digitize data, whether it's born in the cloud or born on premises, this is actually an accelerant to the solutions that we build together. I mean, as blue data, we are all in on the cloud. We support all the three major cloud providers. That was the big announcement that we made this week. We are generally available for AWS, GCP, and Azure. And in particular, we start with customers who weren't born in the cloud. So you're talking about some of the large financial services of all kinds of things. If you have a legacy of all kinds of things. Yeah. You can take that into our account. Yeah, we had Barclays UK here, whom we nominated. They won the Cloudera Data Impact Award. And what they're actually going through right now is they started on prem. They have these really packaged certified technology stacks, whether they are Cloudera Hadoop, whether they are Anaconda for data science. And what they're trying to do right now is, they're obviously getting value from that on-premises with blue data. And now they want to leverage the cloud. They want to be able to extend into the cloud. So we as a company have made our product a hybrid cloud-ready platform. So it can span on-prem as well as multiple clouds. And you have the ability to move the workloads from one to the other, depending on data gravity, SLA considerations. To find city. Yeah. But I think it's one more thing. I want to test this to you guys, John. And that is that analytics is, I don't want to call it inert, but analytics is always, or passive, but analytics has always been about getting the right data to human beings so that they can make decisions. And now we're seeing, because of AI, the distinction that we draw between analytics and AI is, AI is about taking action on the data. It's about having a consequential decision that actually were consequential action as a consequence or as a result of the data. So in many respects, a lot of it's CEO Kubernetes, a lot of these tooling, not only do some interesting things for the infrastructure associated with big data, but they also facilitate the incorporation of new classes of applications that act on behalf of the brand. Here's the other thing I'll add to it. There was a time element here, right? It used to be we were passive, and it was in the past, and you're trying to project forward. That's no longer the case. You can do it right now. Exactly. Yeah, in many respects, the history of the computing industry can be drawn in this way. You're focused on the past, and then with spreadsheets in the 80s of virtual computing, you're focused on getting everybody to agree on the future, and now it's about getting action to happen right now. At the moment it happens. And that's why there's so much action. We're past the setup phase, and I think this is why we're hearing seeing machine learning being so popular because it's like, people want to take action. There's demand. That's a signal that it's time to show where the ROI is and get action done. Clearly, they see that. We're a capitalist, right? And at the end of the day, we're all trying to figure out how to make money in these spaces. Well, certainly there's a lot of movement, and Cloud has proven that spinning up an instance concept has been a great thing, and certainly analytics, it's okay to have these workloads, but how do you tie it together? So, Anotha, I want to ask you, because you guys have been involved in containers, Cloud has certainly been a tailwind. We agree with you 100% on that. What is the relevance of Kubernetes and Istio? You're starting to see these new trends. Kubernetes, Istio, Kubeflow, it's higher level microservices with all kinds of stateful and stateless dynamics. I mean, I call it API 2.0. It's a whole another generation of abstractions that are going on that are creating some goodness for people. What is the impact, in your opinion, of Kubernetes in this new revolution? Yeah, no, I think the impact of Kubernetes is, you know, I was just gave a talk yesterday here. It's called the Hadoopla about Kubernetes. And we're thinking very deeply about this. Yeah, so we're thinking deeply about this. And so I think, you know, Kubernetes, if you look at the genesis, it's all about stateless applications. And I think as new applications are being written, folks are thinking about writing them in a manner that are decomposed, stateless, microservices. You think things like Kubeflow. But, and I think when you write it like that, Kubernetes fits in very well and you get all the benefits of auto scaling and all the sort of the controller pattern, right? Ultimately, Kubernetes is the sort of finite state machine type model where you describe what the state should be and it will work and crank towards making it towards that state. I think it's a little bit harder for stateful applications. And I think that's where we believe that the Kubernetes community has to do a lot more work. And, you know, folks like Blue Data are going to contribute to that work, which is how do you bring stateful applications like, you know, Hadoop, where there's a lot of interdependent services. They're not necessarily microservices. They're actually monolithic applications that are almost close to monolithic application. So I think new applications, new AI, ML tooling that's going to come out, they're going to be very conscious of the how, they're running in a cloud world today that folks weren't aware of seven or eight years ago. So it's really going to make a huge difference. And I think things like Istio are going to make a huge difference because you can start in the cloud and maybe now expand on to on-prem. So there's going to be some interesting dynamics that are going to be connected. Without hop and management frameworks, without hop and, absolutely. Yeah, I mean, this is really critical. I mean, just nailed it. I mean, stateful is where ML will shine if you can then cross the casements of the on-premise where the workloads can have state sharing, scales beautifully, it's a whole nother level. Well, you're going to move the data in the action or the activity, you're going to have to move the processing to the data. And you want to have nonetheless a common seamless management development framework so that you have the choices about where you do those things. Absolutely. Great stuff. We can do a whole CUBE segment just on that. We love talking about these new dynamics going on. We'll see a CNCF, KubeCon coming up in Seattle. Great to have you guys on. Thanks and congratulations on the relationship with Big Blue Data and Dell EMC and Ready Solutions. This is theCUBE with the Ready Solutions here, New York City talking about big data and the impact future of AI, all things stateful, stateless, cloud is driving it all. It's theCUBE bringing you all the action. Stay with us for more after this short break.