 Live from Boston, Massachusetts, extracting the signal from the noise, it's theCUBE covering HP Big Data Conference 2015, brought to you by HP Software. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live in Boston, Massachusetts for HP's Big Data Conference. This is a special presentation of theCUBE, our flagship program where we go out to the events and extract the signal from the noise. I'm John Furrier with Dave Vellante here at Wikibon.org Research. Our next guest, Stephanie McRenals, VP Mark Elation, hot news startup that's been kind of coming out of stealth, it's out there, Big Data, a lot of great stuff. Stephanie, welcome to theCUBE. Great to see you. Great to be here. Tell us about the startup first of all because good buzz going on, it's kind of stealth buzz but it's really with the thought leaders and really the people in the industry who know what they're talking about, like what you guys are doing. So introduce the company and tell us what you guys are doing and relationship with Vertica and exciting stuff. Absolutely, Elation is an exciting company. We just started to come out of stealth in March of this year. We came out of stealth with some great production customers, so eBay is a customer. They have hundreds of analysts using our system. We also have Square as a customer, smaller analytics team but the value that analytics teams are getting out of this product is really being able to access their data in human context. So we do some machine learning to look at how individuals are using data in an organization and take that machine learning and also gather some of the human insights about how that data is being used by experts, surface that all, in line with the work plans. So what kind of data? Because Stonebreaker was kind of talking yesterday about the three Vs which we all know but the one that's really coming mainstream in terms of problem space is variety. Variety, you have the different variety of schema sources and then you have a lot of unstructured exhaust or data flying around. Can you be specific on what you guys do? Yeah, I mean it's interesting because there's several definitions of data and big data going around. And so we connect to a lot of database systems and we also connect to a lot of Hadoop implementations. So we deal with both structured data as well as what I consider unstructured data. And I think the third part of what we do is bringing context from human created data or human information with which Robert yesterday was talking about a little bit which is, what happens in a lot of analytic organizations is that there's a very manual process of documenting some of the data that's being used in these projects and that's done on wiki pages or spreadsheets that are floating around the organization and that's actually a really rich. Slack, Basecamp, all these collaboration. All these collaboration platforms and what you realize when you start to really get into the work of using that information to try to write your queries is that trying to reference a wiki page and then write your sequel and flip back and forth between maybe 10 different documents is not very productive for the analysts. So what our customers are seeing is that by consolidating all of that data and information in one place where the tables are actually referenced side by side with the annotations, their analysts can get from 20 to 50% savings in productivity and new analysts, maybe more importantly new analysts can get up to speed quite a bit quicker. At Square Day I was talking to one of the data scientists and he was talking about his process for finding data in the organization, which prior to using Alation, it would take him about 30 minutes going to maybe three or four people to find the data he needed for his analysis. And with Alation in five seconds, he can run a query, search for the data he wants, gets it back, gets all that expert annotation already around that base data set and he's ready to roll. He can start testing some of this hypothesis. So you call it a platform, right? You've heard it was a platform and you said you work with a lot of databases, right? So it's tightly integrated with the database in this use case? So it's interesting, you know, we see databases as a source of information so we don't create copies of the data on our platform. We go out and point to the data where it lies and surface that data to the end user. Now in the case of Vertica and our relationship with Vertica, we've also integrated Vertica into our stack to support what we call data forensics, which is the ability for not an analyst who's using the system day to day, but for an IT individual to understand what are the behaviors around this data and the types of analysis that are being done. And so Vertica is a great high-performance platform for dashboarding and business intelligence, the backend of that, providing quick access to aggregates. So one of the things we do- Well what part of Vertica are you guys? Just the engine, what specifically again? Yeah, so we use the Vertica engine underneath our forensics product and then that's one portion of our platform, the rest of our platform is built out on other technologies. So Vertica's part of your solution, is that right? It's part of our solution. It's one application that we, part of one application that we deliver. So we've been talking all week about this, when Colin Mahoney in his talk yesterday, I don't know if you saw it, but he gave a little history on ERP, how initially it was highly customized and it became packaged apps. And he sort of pointed to a similar track with analytics. Although he said it's not going to be the same, it's going to be more composable sort of applications. I wonder, and historically the analytics in the database have been closely aligned, I'll say, maybe not integrated. Do you see that model continuing? Do you see it more packaged apps or more what Colin's calling composable apps? What's the relationship between your platform and the application layer? Yeah, so our platform is really more tooling for those individuals that are building or creating those applications. So we're helping data scientists and analysts find what algorithms they want to use as a foundation for those applications. So a little bit more on the discovery side, where folks are doing a lot of experiment experimentation. They may be having to prepare data in different ways in order to figure out what might work for those applications. And that's where we fit in as a vendor. And what's your license model? And so we're on a subscription model. We have customers that have data teams in the hundreds at a place like eBay. The smaller implementations could be maybe just teams of five analysts, 10 analysts, fairly small rates. So it's a seat-based subscription? It's a seat-based subscription, but we can run in the cloud, we can run on-premise. We do some interesting things around securing the data where you can see your columns but not the data sets for financial services organizations and our customers that have security concerns. And most of those are on-premise implementations. Stephanie, talk about the inspiration of the company and about the company. It's been three years since they came out of stealth. What's the founders like? What's the DNA of the company? What do you guys do differently and what was the inspiration behind this? Yeah, what's really interesting, I think about the founding of the company is that the technical founders come from both Google and Apple. So you have an interesting observation that both individuals had made independently about. Hardcore algorithmic guy and then like relevant, clean. Yeah, well, and both folks kind of made interesting observations about how Google and Apple, two of the most data-driven companies on the planet, were struggling and their analytics teams were struggling with being able to share queries and share data sets. And there was a lot of replication of work that was happening. And so both of these folks from different angles kind of came together at Elation and said, look, there's a lot of machine learning algorithms that could help with this process. And there's also a lot of good ways with natural language processing to let people interact with their data in more natural ways. The founder from Apple, Aaron, he was on the Siri team. So he had a lot of experience designing products for navigability and ease of use and natural language learning. And so those two perspectives coming together have created some technology fundamentals in our product that's pretty exciting. And there's, too, some scar tissue from large-scale implementations of data. Yeah, very large-scale implementations of data and also a really deep awareness of what the human equation brings to the table. So machine learning algorithms aren't enough in and of themselves. And I think Ken Rudin had some interesting comments this morning where he kind of pushed it one step further and said, it's not just about finding insight. Data science is about having impact and you can't have impact unless you create human context and you have communication and collaboration around the data. So we give analysts a query tool by which we surface the machine learning context that we have about the data that's being used in the organization and what queries have been running that data. But we surface it in a way where the human can get recommendations about how to improve their sequel and drive towards impact and then share that understanding with other analysts in the organization so you get an innovation community that's started. So who are you guys targeting? Let's step back and go to market now. You guys are launched. Got some funding. Can you share the amount? Or is it private, confidential? Or how much did you raise? Who are you targeting? What's your go-to-market? What's the value proposition? Give us the data. Yeah, so the initial value proposition is just really about analyst productivity. That's where we're targeted. How can you take your teams of analysts? And everyone knows it's hard to hire these days. So you're not going to be able to grow those teams out overnight. How do you make the analysts, the data scientists, the PhDs you have on staff much more productive. How do you take that 80 to 90% of the time that they're- Get them using stuff, sharing data. Get them using stuff, get them sharing data. Try to get them out of the tedium of trying to just find data in the organization and prepare it and let them really innovate and use that to drive value back to the organization. So we're often selling to individual analysts, to analytics teams. The go-to-market starts there and the value proposition really extends much further in the organization. So you find teams and organizations that have been trying to document their data through traditional data governance means or ETL tools for a very long time and a lot of those projects have stalled out. The way that we crawl systems and use machine learning automation to automate some of that documentation really gives those projects a new life in a lot of organizations. And the price data has always been elusive. I mean, did we go back decades, structured data, all these pre-built databases, it's been hard, right? So if you can crack that nut, that's going to be a very lucrative, and there's opportunity now, you got to do clusters now, storing everything. I mean, some clients we talked to here on theCUBE or customers of ASA, HP or IBM and big companies, they're storing everything just because they don't know what to do with it yet. Yeah, I mean, the past has been hard in part because we, in some cases, over-manage the modeling of the data. And I think what's exciting now about storing all your data in Hadoop and storing first and then asking questions later is you're able to take a more discovery-oriented, hypothesis-testing, iterative approach. And if you think about how true innovation works, you build insights on top of one another to get to the big breakthrough concepts. And so I think we're at an interesting point in the market for a solution like this that can help with that increasing complexity of data environment. So you just raised your Series A, raised 9 million, you maybe did some seed round before that. So pretty early days for you guys. You mentioned natural language processing before, one of your founders. Are you using NLP in your solution in any way? So we have a search interface that allows you to look for that technical data, to look for metadata and for data objects by entering a simple natural language search term. So we are using that as part of our interface and solution. Right, and so early customer successes, can you talk about any examples? Yeah, there's some great examples. Jointly with Vertica, Square is a customer and their analytics team is using us on a day-to-day basis not only to find data sets in the organization but to document those data sets. eBay has hundreds of analysts that are using Alation Today in a day-to-day manner. They've seen quite a bit of productivity out of their new analysts that are coming on the systems that used to take analysts about 18 months to really get their feet around them in the eBay environment because of the complexity of all of the different systems that eBay and understanding where to go for that customer table that they needed to use. Now analysts are up and running about six months and their data governance team has found that Alation has really automated and prioritized the process around documentation for them and so it's a great, late a great foundation for then their data curators and data stewards to go in and enrich the data and collaborate more with the analysts and the actual data users to get to a point of cataloged data that's useful to the audience. So what's next? You guys are going to be on the road in New York, post-tractic Hadoop world, big data NYC is coming up, the big event in New York, theCUBE will be there. We're getting the word out about Alation and what we're doing. We have customers that are starting to speak about their use cases and the value that they're seeing. We'll be in New York, market share I believe, we'll be speaking on our behalf there to share their stories and then we're also going to a couple other conferences after that. You know, the fall is an exciting time for- Which one's your big ones there? So we'll be at Strata in New York and to September, early October and then mid-October we're going to be at both Teradata Partners and Tableau's Conference as well. So we connect not only to databases of all different sorts, but also to the intelligence tools. Yeah, awesome. Well, anything else you'd like to add, share with the company is awesome. We've heard some great things about you guys, been checking around. I've seen, found out about you guys and a lot of people like the company. I mean, a lot of insiders like, you didn't raise too much cash, just raised a lot and that's not the million, zillion dollar round. I think what, did you raise like nine million? Yeah, we raised nine million and I think we're building this company in a traditional value-oriented way. Ray, we're- Hey, what a great idea. We're tracking revenue. Hey, stay alive. We're bringing in revenue and trying to balance that out with venture capital investment and it's not that we won't take money, but we want to build this company in a very value-oriented way. Yes, a durable, so the vision is to build a durable company. Absolutely, absolutely. And that may be different than some of our competitors out there these days, but that's the path we think- Dave and I have not taken any financing on SiliconANGLE at all, so again, we believe in that and you might pass up some things but you know what, have control and you guys have some good partners. So congratulations. Thank you. Final word, what's this conference like? You go to a lot of events. What's your take on this event? Yeah, I do end up going to a lot of events as part of the marketing role. You know, I think what's interesting about this conference is that there are a lot of great conversations that are happening and happening not just from a technology perspective, but also between business people and deep thinking about how to innovate and Vertica's customers I think are some of the most loyal customers I've seen in the market, so it's great to see- And they're advanced too. They're talking about some pretty big problems that they're solving. It's not like little point solutions, it's more re-architecting. Some DevOps, I got a DevOps, I got trashed on Twitter, private messages all last night about me calling this a DevOps show. It's not really a DevOps Cloud show but there's a DevOps vibe here, the people who are working on these solutions. I think they're just a real vibe. People are solving real problems and they're talking about them and they're sharing their opinions and I think that's similar to what you see in DevOps. The guys in DevOps are in the front lines. They're real engineers. They're engineering stuff. Yeah, they have to engineer because of the pressures there are other. No pretenders here, that's for sure. We were talking earlier, it's not a big sales conference, right? It's a lot of customer content. They're engineering solutioners talking to peers. They don't want the bullshit. They don't want real. I mean, I got a lot on the table. I'm doing some serious work and I want serious conversations and that's refreshing for us. I mean, we love, love events like this. All right, Stephanie, thanks for so much for coming on theCUBE, sharing your insight. Congratulations, good luck with the new startup. Hot startups here in Boston here at the Vertica HP Software Show. We'll be right back more on theCUBE after this short break.