 from San Jose in the heart of Silicon Valley. It's theCUBE, covering DataWorks Summit 2018. Brought to you by Hortonworks. Welcome back to theCUBE's live coverage of DataWorks here in San Jose, California. I'm your host, Rebecca Knight, along with my co-host, James Kobielus. We have two guests in this segment. We have Steve Woolwich. He is the VP of product marketing at Arcadia Data and Satya Ramachandran, who is the VP of engineering at Neustar. Thanks so much for coming on theCUBE. Our pleasure and thank you. So let's start up by setting the scene for our viewers. Tell us a little bit about what Arcadia data does. Arcadia data is focused on getting business value from these modern scale-out architectures like Hadoop and the Cloud. We started in 2012 to solve the problem of how do we get value into the hands of the business analysts that understand a little bit more about the business in addition to empowering the data scientists to deploy their models and value to a much broader audience. I think that's been, in some ways, the last mile of value that people will need to get out of Hadoop and data lakes is to get it into the hands of the business. So that's what we're focused on. And start seeing the value, as you said. Yeah, seeing is believing. Pictures is a thousand words, all those good things. And what's really emerging, I think, is companies are realizing that traditional BI technology won't solve the scale and user concurrency issues because architecturally, big data's different, right? We're on these scale-out NPP architectures now like Hadoop, the data complexity and variety's changed, but the BI tools are still the same and you pull data out of the system to put it in some little micro-cube and do some analysis. Companies want to go after all the data and do the analysis across a much broader set and that's really what we enable. And I want to hear about the relationship between your two companies, but Satya, tell us a little bit about NuStar, what you do. NuStar is an information services company. We have built around identity. We are the premier identity provider, the most authoritative identity provider for US and we've built a whole bunch of services around that, the identity platform. I am part of the marketing solutions group. I head the analytics engineering for marketing solutions. And the product that I work on helps marketers do their annual planning as well as their campaign or tactical planning so that they can fine tune their campaigns on an ongoing basis. And how, so how do you use Arcadia data's primary product? So we are a predictive analytics platform. The reporting solution, we use Arcadia for the reporting part of it. So we have multi-terabytes of advertising data in our warehouse and so we use Arcadia to provide fast access to our customers and also very granular and explorative analysis of this data, iterative and explorative analysis of this data. So you say you help your customers with their marketing campaigns. So are you doing predictive analytics? Correct. And are you doing churn analysis and so forth and how does Arcadia fit into all of that? So we get data and then we build an attribution model which tells how the marketing spend corresponds to the revenue. We not only do historical analysis, we also do predictive in the sense that the marketeers frequently run water analysis saying that what if I moved my budget from freight search to TV and how does it affect the revenue? So all of this modeling is built by Neustard. The modeling platform is built by Neustard. But the last mile of taking these reports and providing this explorative analysis of the results that is provided by the reporting solution which is Arcadia. Well, I mean the thing about data analytics is that it really is going to revolutionize marketing. That famous marketing adage of I know my advertising works I just don't know which happened. Now we're really going to be able to figure out which half, can you talk a little bit about return on investment and what your clients see? Sure, we've got some major Fortune 500 companies that have said publicly that they've realized over a billion dollars of incremental value. And that could be across both marketing analytics and how do we better treat our messaging and brand and reach our intended audience. It's things like supply chain and being able to more real time analyze what if analysis for different routes. It's things like cybersecurity and stopping fraud and waste and things like that at a much grander scale than what was really possible in the past. So we're here at DataWorks and it's the Hortonworks show. Give us a sense for the degree of your engagement or partnership with Hortonworks and participation in their partner ecosystem. Yeah, absolutely. Hortonworks is one of our key partners and what we did that's different architecturally is we built our BI server directly into the data platform. So what I mean by that is we take the concept of BI server we install it and run it on the data nodes of Hortonworks data platform. We inherit the security directly out of systems like Apache Ranger. So that all that administration and scale is done at Hadoop economics, if you will. And it leverages the things that are already in place. So that has huge advantages both in terms of scale but also simplicity. And then you get the performance of the concurrency that companies need to deploy out to like 5,000 users directly on that Hadoop cluster. So Hortonworks is a fantastic partner for us and a large number of our customers run on Hortonworks as well as other platforms such as Amazon Web Services where Satya's got his system deployed. At this show they announced Hortonworks data platform 3.0 there's containerization in there there's updates to Hive to enable it to be more of a real-time analytics and also a data warehousing engine. And Arcadia data do you follow their product enhancements in terms of your own product roadmap with any specific fixed cycle? Are you going to be leveraging the new features in HDP 3.0 going forward to add value to your customer's ability to do interactive analysis of this data in close to real-time? Yeah, because we're a native. The campaigns are often in real-time increasing especially when you're doing all your, you get a completely digital business. Yeah, absolutely. So we benefit from the innovations happening within the Hortonworks data platform. So because we're a native BI tool that runs directly within that system with changes in Hive or different things within HDFS in terms of performance or compression and things like that our customers jointly benefit from that directly. So yeah. Satya, going forward, what are some of the problems that you want to solve for your clients? I mean, what are their biggest pain points and where do you see new stars as? So data is the new oil, right? So marketeers also, for them, now data is what they're going after. They want faster analysis, they want to be able to get to insights as fast as they can and they want to obviously get work on as large amount of data as possible. The variety of sources is becoming higher and higher and higher in terms of marketing. There used to be a few channels in 70s and 80s and 90s kind of increased it. Now you have like hundreds of channels, if not thousands of channels and they want visibility across all of that. It's the ability to work across this variety of data increasing volume at a very high speed that those are high level challenges that we want. So there was marketing attribution analysis. You say it's one of the core applications of your solution portfolio. How is that more challenging now than it had been in the past? We have far more marketing channels, digital and so forth. And then how does the state of the art of marketing attribution analysis, how is it changing to address this multiplicity of channels and media for advertising and for influencing the customer on social media and so forth? And then, you know, it gives us a sense of what are the necessary analytical tools needed for that? We often hear about social graph analysis for semantic analysis or for behavioral analytics and so forth. All of this makes very challenging. How can you determine exactly what influence is a customer now in this day and age where Twitter is an influencer over the conversation? How can you nail that down to specific KPIs or specific things to track? So I think from our, like you pointed out, the variety is increasing, right? And I think that the marketing has now have a lot more options than what they have and that's a blessing and now it's also a curse because then I don't know where I'm kind of moving my marketing spend to. So I was just, so attribution right now is still sitting at the headquarters, right? It's kind of sitting at a very high level and it is answering questions. Like we sit with the 1,400 companies, it's still answering questions to the CMOs, right? Where attribution will take its next step is to then move lower down where it's able to answer the regional headquarters on what needs to happen and more importantly on every store I'm able to then answer and tailor my attribution model to a particular store. Let's take Ford for an example, right now we sit at the CMO suite but if I'm able to go to every dealer and I'm able to personalize my attribution model to that particular dealer, then it becomes a lot more useful. The challenge there is it all needs to be connected. Whatever model we are working for the dealer needs to be connected up to the headquarters. And then personalization, it very much leverages the kind of things that Steve was talking about with Arcadia, being able to analyze all the data to find those micro, micro, micro segments that can be influenced to varying degrees. So yeah, I like where you're going with this because that very much relates to the power of distributed federated big data fabrics like what Hortonworks offers, yeah. Correct and so it's streaming analytics is coming to four and it's been talked about for the longest period of time but we have real use cases for streaming analytics right now. Similarly, the large volumes of the data volume is indeed becoming a lot more. So both of them are becoming a lot more. Yes. Great. Well Satya and Steve, thank you so much for coming on theCUBE. This was really, really fun talking to you. Thanks, absolutely. Great to meet you. Thanks for having us. Yeah, it's fun. I'm Rebecca Knight for James Kobielus. Stay tuned to theCUBE. We will have more coming up from our live coverage of DataWorks just after this.