 Hi, I'm Peter Burris, and welcome to another Cube Conversation from our wonderful studios in Palo Alto, California. Great conversation today. We've got Christian Radatas, who's the CEO of Datamere here to talk about some of the trends within the overall analytics space, one of the most important things happening in technology today. Christian, welcome back to theCUBE. Good morning, Peter. Thanks for having me today. Oh, it's great to have you here. Hey, let's start with kind of some of the preliminaries. What's happening at Datamere? Well, we've been around for nine years now, which is a lot of time in a very agile technology space. And I actually just came back from an investing offsite that was arranged from one of our biggest investors and everything is centering around the cloud. And we were trotting along within the Hadoop ecosystem, the big data ecosystem over the past couple of years. And since 12, 15 months, the transition in the analytics market and how it's transforming from on-premise to the cloud in a hybrid way as well as has been stunning, right? And we are faced with a challenge in innovating in those spaces and making our product relevant for on-premise deployments, for cloud deployments and various different cloud platforms and in a hybrid fashion as well. And we've been traditionally working with customers that have been laggards in terms of cloud adoption because we do a lot of business and financial services and insurance, healthcare, telecommunications. But even in those industries over the past year, it has been stunning how they accelerate cloud adoption and how they move analytic workloads to the cloud. Well, interesting, those sound like sometimes leaders in the analytics world, even if they're laggards in the cloud and there's something of a relationship there where people didn't want to do a lot of their analytics because they were doing analytics in some of the most strategic sensitive data and they felt the pressure to not give that off to a company that they felt perhaps or an industry that's a little bit less ready from an infrastructure standpoint. But our research shows pretty strongly that we're seeing a push to adoption precisely because so much of that ecosystem got wrapped up in the infrastructure and never got to the possible value of analytics. So is that helping to force this along, do you think? The idea of- Absolutely, right? If you look at the key drivers and there was some other analyst research that I read this week, why are people being motivated, moving the analytic workloads into the cloud? It's really less cost, it's really business agility. How do they become independent from IT and procure services across the organization in a very simple, easy and fast fashion? And then there's a lot of fears associated with it, right? It's data governance, it's security, it's data privacy is what these industries that we predominantly work in are concerned with, right? And we provide a solution framework that actually helps them to transition those on-premise analytic workloads into the cloud and still get the enterprise-grade features that they're used to from an on-premise solution deployment. Yeah, so in other words, a lot of businesses confused failure to deal with big data infrastructure as failure to do big data. That's correct. I want to build on something you just said, specifically the governance issue because I think you're absolutely right. There's an enormous lack of understanding about what really constitutes data governance. It used to be, oh, data governance is what the data administrator does when they do modeling and who gets to change the model and who owns the model and who gets to you, all that other stuff. We're talking about something fundamentally different as we embed more deeply some of these analytics directly into high-value business activities that are being utilized or performed by high-cost business executives. Absolutely. How does data governance play out? And I'm going to ask you specifically, what are you guys doing that makes data governance more accessible and more manageable within Datamere customers? So I think there's two key features to our solution that support this. So number one, we have very much a self-service aspect to it. So we're pushing the abilities to model and create views on the big data assets that are persisting in the data lakes towards the business user, right? But we do this in a very governed way, right? We can provide full data lineage. So we can audit every single step how the data is being sourced, how it's been manipulated on the way and provide an audit trail, which is very important for many of the customers that we work with. And we really bring this into the hands of the business users without much IT interference. They don't have to work on models to be built and so on and so forth. And this is really what helps them build rapid analytic applications that provide a lot of value and benefits for their business processes. So you talked about how you're using governance or the ability to provide a manageable governance regime to open up the aperture on the utilization of some of these high value analytics frameworks by broader numbers of individuals within the organization. That seems to me to be a pretty significant challenge for a lot of businesses. It's not enough to just have a ivory tower group of data scientists be good at crafting data, understanding data, and then advising people what actions to take based on that data. It seems it has to be more broadly diffused within an organization. What do you think? So this is clearly the trend and as these analytic services move to the cloud, you will see this even more so, right? You will have curated data assets and you provide access control for certain user groups that can see and work with this data, but then you need to provide a solution framework that enables these customers to consume this in a very seamless and easy way. And this is basically what we are doing. We want to push it down to the end user and give them the ability to work on complex analytical problems using our framework in a governed way, in a fast way, in a very iterative analytic workflow. A lot of our customers, they have analytic or they pursue analytic problems that are of investigative nature. In this, you cannot do this if you rely on IT to build new models to delay the process. Or if you only rely on IT. Or if you only rely on IT, right? They want to do this on their own and create their own views depending on the analytical workflow in a very rapid way. And so we support this in a highly governed way. They can do this in a very fast and rapid fashion. And as it moves to the cloud, it provides some of the even more opportunities to do so. So as CEO of Datamir, you're spending a lot of time with customers. Are there some patterns that you're seeing customers in addition to buying Datamir? But are there some patterns in addition to what you just described that the successful companies are utilizing to facilitate this diffusion? Are they training people more? Are they embedding this more deeply into other types of applications or workloads? What are some of those patterns of success that you're seeing amongst your customers? So this is a very interesting question, right? Because a lot of big data initiatives within companies fail for the lack of adoption, right? So they build these big data lakes or ramp up cloud services and they never really see adoption. And so the successful customers we work with, they have a couple of things they do differently than others. So they have a centralized COE type of organization usually that facilitates and promotes and educates people on number one, the data assets that are being available through the organization, about the tool sets that are being used amongst one of them, obviously, Datamir within our customers. And they facilitate constant education and experience sharing across the organization for the use of big data assets throughout the organization. And these companies, they see adoption, right? And it spreads throughout the organization and becomes an increasing momentum and adoption across various business departments for many high value use cases. So we've done a lot of research. I myself have spent a lot of time in questions of technology adoption, questions within large enterprises. And you absolutely described it, it fails to adopt. From an adoption standpoint, it's called they abandon. Absolutely true. One of the things that often catalyzes whether or not someone continues to adopt or a group determines to abandon is a lack of understanding of what the returns are, what kind of returns these changes of behavior are initiating or instantiating. And I've always been curious why a lot of these software tools don't do a good job of actually utilizing data about utilization from a big data standpoint to improve the adoption of big data. Are you seeing any effort made by companies to use Datamir to help businesses better adopt Datamir? I haven't seen that yet. I see this more with our OEM customers. So we've got OEM customers that analyze the cloud consumption with their customers and provide analytics on usage across the organizations. I see these things. And from our standpoint, we facilitate this process by providing use case discovery workshops. So we have a services organization that helps our customers to see the light literally to understand what's the nature of the data assets available. How can they leverage for a specific use case, high value use case, implementations, experience sharing, what are other customers doing? What kind of high value applications are they going after in a specific industry and things like this? We do lunch and learns with our customers. We just recently did one with a big healthcare provider and the interest is definitely there. You get 200 people in a room for a lunch and learn meeting and everybody's interesting how they can make their life easier and make better business decisions based on the data assets that are available throughout the organization. Yeah, it's amazing when a lunch and learn meeting goes from 20 people to 200 people that really becomes much more focused on learning. So one of the question I have related to this is that you've got a lot of experience in the analytics space, more than big data and how the overall analytics space has evolved over the years. We have some research that pretty strongly suggests that it's time to start thinking about big data not as a thing unto itself, but as part of an aggregate approach to how enterprises should think about analytics. What do you think, how do you think an enterprise should start to refashion its understanding of the role that big data plays in a broader understanding of analytics? So if you, if I look back in the earlier days of my career, I come from the EDW world, right? And then you had, all the large enterprises had EDWs and they tried to build this centralized repository of data assets. Highly modeled. Highly modeled. A lot of work to set up, structured, highly modeled, extreme complex to modify and service a new application request from business users and then came the Hadoop data lake-based big data approach that said dumped the data in. And this is where we were apart within where we became very successful in providing a tool framework that allows customers to build virtual views into these data assets in a very rapid fashion driven by the business user community. But to some extent these data lakes have also had issues in servicing the bread and butter BI user community throughout the organization. And the EDW never really went away, right? So now we have EDWs, we have data lakes that service different analytic application requirements throughout the organization. And even reporting systems. And even reporting systems. And now the third wave is coming by moving workloads into the cloud. And if you look into the cloud, the wealth of available solutions to a customer becomes even more complex. The cloud vendors themselves build out tons of different solutions to service different analytical needs, the marketplaces of hundreds of solutions of third party vendors. And the customers try to figure out how all these things can be stitched together and provide the right services for the right business user community throughout the organization. So what we see moving forward will be a hybrid approach that will retain some of the on-premise EDW and data lake services. And those will be combined with multi-cloud services. There will also not be a single cloud service. And we are already seeing this today. So one of our customers is Sprint Pinsight, the advertising business of the Sprint Telecommunications companies. They have a massive Hadoop on-premise data lake. And then they do all the pre-processing of the ads data from the network with data on-premise. And we condense down the data asset from a daily volume of 70 terabytes to eight. And this gets exposed to a SQL cloud-based data warehouse service for BI consumption throughout the organization. So you see these hybrid, very agile services emerging throughout our customer base. And I believe this will be the future. Yeah, one of the things we like about the concept or the approach of virtual view is precisely that. It focuses in on the value that the data is creating and not the underlying implementation so that you have greater flexibility about whether you treat it as a big data approach or an EW approach or whether you put it here or whether you put it there. But by focusing on the outcome that it's delivered, it allows you a lot of flexibility in the implementation you employed. I agree. Phenomenal, Christian Rodadis, CEO of Datamirth. Thanks again for being on theCUBE. Appreciate it, thanks Peter. You bet.