 From theCUBE Studios in Palo Alto in Boston, connecting with thought leaders all around the world, this is a CUBE Conversation. Hey, welcome back, everybody. Jeff Frick here with theCUBE. We're coming to you today from our Palo Alto studios with a CUBE Conversation, talking about data. And we're excited to have our next guest. He's been on a number of times, many times CUBE alum, really at the forefront of helping companies and customers be more data-centric in their activities. So we'd like to welcome on to the show Satyan Sangani, he is the co-founder and CEO of Elation. Satyan, great to see you. Great to see you, Jeff. It's good to see you again in this new world, a new format. It is a new world, a new format. And what's crazy is, you know, in March and April, we were talking about this light switch moment and now we've just turned the calendar to October and it seems like we're going to be doing this thing for a little bit longer. So, you know, it is kind of the new normal. And even I think when it's over, I don't think everything's going to go back to the way it was. So, so here we are. But you guys have some exciting news to announce. Let's just jump to the news and then we'll get into a little bit more of the nitty gritty. So what do you got coming out today, right? What we are announcing today is basically Elation 2020, which is probably one of the biggest releases that I've been with, that we've had since I've been with the company. We, with it, are releasing three things. So in some sense, there's a lot of simplicity to the release. The first thing that we're releasing is a new experience around what we call the business user experience, which will bring in a whole new set of users into the catalog. The second thing that we're announcing is basically around Elation analytics. And the third is around what we would describe as a cloud native architecture. You know, in total, it brings a fully transformative experience, basically lowering the total cost of getting to a data management experience, lower and a data intelligent experience, much lower than previously had been the case. And you guys have a really simple mission, right? You're just trying to help your customers be more data, what's the right word? You know, data-centric, use data more often, you know, and to help people actually make that decision. And you had an interesting quote in another interview you talked about trying to be the yelp for information, which is such a nice kind of humanizing way to think about it, because data isn't necessarily that way. And I think, you know, you mentioned before we turn on the cameras that for a lot of people, maybe it's just easier to ignore the data if I can just get the decision through on gut and intuition and get on to my next decision. Yeah, you know, it's funny. I mean, we live in a time where people talk a lot about fake news and alternative facts. And our vision is to empower a curious and rational world. And, you know, I always smile when I say that a little bit because it's such a crazy vision, right? Like, how do you get people to be curious and how do you get people to think rationally? But, you know, to us, it's about one, making the data really accessible, just allowing people to find the data they need when and as they want it. And the second is for people to be able to think scientifically, you know, teaching people to take the facts at their disposal and interpret them correctly. And, you know, we think that if those two skills existed, just the ability to find information and interpret it correctly, people could make a lot better decisions. And so the yelp analogy is a perfect one because if you think about it, yelp did that for local businesses just like Amazon did it for really complicated products on the web. And what we're trying to do at Elation is in some sense, very simple, which is to just take information and make it super usable for people who want to use it. Great. But I'm sure there's the critics out there, right? Who say, yeah, we've heard this, you know, before the promise of BI has been around forever. And I think a lot of people's thinking it just didn't work, whether the data was too hard to get access to, whether it was too hard to manipulate, whether it was too hard to pull insights out, whether it's just too much scrubbing and manipulating. So, you know, what is some of the secret sauce to take what is a very complex world? And again, and you got some very large customers with some giant data sets and to, I don't want to say humanize it, but kind of humanize it and make it easier and more accessible for that business analyst, not just generally, but more specifically when I need it to make a decision. Yeah, I mean, it's so funny because, you know, making something, data is like a lot of software and death by a thousand cuts. I mean, you know, you look at something from the outside and it looks really, really, really simple, but then you kind of dwell into any problem and that can be CRM, you know, something like Salesforce or it can be something like ServiceNow with, you know, ITSM, but these are all really, really complicated spaces and getting into the depths and the detail of it is really hard. And data is really no different, right? Data is just the sort of exhaust from all of those different systems that exist inside of your company. So the detail around the data in your company is exhaustingly, you know, minute. And so, you know, how do you make something like that simple? You know, I think really the biggest challenge there is progressively revealing complexity, right? Giving people the right amount of information at the right amount of time. So one of the really clever things that we do in this business user experience is we allow people to search for and receive the information that's most relevant to them. And we determined that relevance based upon the other people in the enterprise that happen to be using that data. And we know what other people are using in that company because we look at the logs to understand which data sources are used most often and which reports are used most often. So right off the bat, when you get something, you just see the name of a report and it could be around the revenues of a certain product line. But the first thing that you see is who else uses it? And that's something that people can identify with. You may not necessarily know what the algorithm was or what the formula might be or how the business glossary term relates to some data model or data artifact, but you know the person. And if you know the person, then you can trust the information. And so, you know, a lot of what we do is spend time on design to think about what is it that a person expects to see and how do they verify what's true? And that's what helps us really understand what to serve up to somebody so that they can navigate this really complicated world of data. That's awesome. Because it's really a signal-to-noise problem, right? And I think I've heard you speak before and of course this is not new information, right? There's just so much data, right? The increasing proliferation of data and it's not that there's that much more data. We're just capturing a lot more of it. So your signal-to-noise problem just gets worse and worse and worse. And so what you're talking about is really kind of helping filter that down to get through a lot of that noise so that you can find the piece of information within the giant haystack that is what you're looking for at this particular time in this particular moment. Yeah. And it's a really tough problem. I mean, one of the things that, you know, it's true that we've been talking about this problem for such a long time. And, you know, in some sense, if we're lucky, we're going to be talking about it for a lot longer. Because, you know, it used to be that the problem was, you know, back when, you know, I was growing up, you know, you were doing research on the topic and you'd go to the card catalog and you'd go to the Dewey Decimal System and in your, you know, elementary school or high school library, you might be lucky if you were to find one, two, or three books that map to the topic that you were looking for. Now you go to Google and you find 10,000 books. Now you go inside of an enterprise and you find 4,000 relational database tables and 200 reports about an artifact that you happen to be looking for. And so really the problem is, what do I trust? And what's correct? And, you know, getting to that level of, you know, accuracy around information, if there's so much information out there is really the big problem of our time. And I think, you know, for me, it's a real privilege to be able to work on it because I think if we can teach people to use information better and better, then they can make better, better decisions and, you know, that can help the world in so many different ways. Right, right. My other favorite example that everybody knows is photographs, right? Back when you only got 24 in a roll and it cost you six bucks to develop it, you know, those were pretty special. Well, now you go buy a fancy camera, you can shoot 11 frames a second, you go out and shoot the kids at the soccer game, you come home with 5,000 photos, how do you find the good photo? It's a real problem. If you've ever faced something like that, there's kind of a splash of water in the face, like where do I even begin? But the other piece that you talk about a lot, which is slightly different, but related is context. You know, and my favorite context, it's like 55, right, that's a number. But if you don't have any context for that number, is it a temperature, is it cold inside the building, is it a speed, is it too slow on I-5, or is it fast because I'm on a bicycle going down a hill? And without context, data is just a number, it doesn't mean anything. So you guys really by adding this metadata around the data are adding a lot more contextual information to help figure out kind of what that signal is from the noise. Yeah, you'll get facts from anywhere, right? Like, you know, you can have the Hitchcock, you can have 55 or 42 and you can figure out like what the meaning of the universe is and apparently the answer is 42 and what does that mean, you know, might mean a million different things. And that, to me, that context is the difference between, you know, suspecting and knowing, and it's the difference between having confidence and basically guessing. And, you know, I think to the extent that we can provide more of that over time, you know, that's what's going to make us, you know, an ever more valuable partner to the customers that we, you know, satisfy today. Right, well, I do know why 42 is always the answer because that's Ronnie Law and that's always the answer. So that one I know, that's an easy one. But it is really interesting. And then you guys just came out, I had Aaron Calbonne, one of your co-founders, the other day and we talked about this new report that you guys have sponsored, the data culture report and really, you know, putting some granularity on a data culture index. And I thought it's pretty interesting and I'm psyched that you guys are going to be doing this longitudinally because whether you do or do not necessarily agree with the method, it does give you a number, it does give you a score, it's a relatively simple formula. And at least you can compare yourself over time to see how you're tracking. I wonder if you could share, I mean, the thing that jumps out right off the top of that report is something we were talking about before we turn the cameras on that, you know, people's perception of where they are on this path doesn't necessarily map out when you go bottoms up and add the score versus top down when I'm just making an assessment. Yeah, it's funny, it's kind of the equivalent of everybody thinks they're an above average driver or you know, everybody thinks they're above average in terms of obviously intelligence. And you know, obviously that mathematically is not possible or true, but I think in the world of data management, you know, we all talk about data, we all talk about how important it is to use data and if you're a data management professional, you want people in your company to use more data, but ironically the discipline of data management doesn't actually use a lot of data itself. It tends to be a very, you know, slow, methodical, process driven, gut oriented process to develop things like what data models exist and how do I use my infrastructure and you know, where do I put my data and which data quality is best? Like, you know, all of those things tend to be, you know, somewhat heuristic driven or gut driven and they don't have to be. And a big part of our release actually is around this product called Elation Analytics. And what we do with that product is really quite interesting. We start measuring elements of how your organization uses data by team, by data source, by use case, and we give you transparency into what's going on with the data inside of your landscape and ecosystem. So you can start to actually score yourself both, you know, internally, but also, you know, as we reveal in our customer success methodology against other customers to understand what it is that you're doing well and what it is that you're doing badly. And so, you know, you don't need necessarily to have a gust instinct anymore. You can look at the data of yourselves and others to figure out where you need to improve. And so that's a pretty exciting thing. And I think this notion that says, look, you know, you think you're good, but are you really good? I mean, you know, that's fundamental to improvement in business process and improvement in data management and improvement in data culture fundamentally for every company that we work with. Right, right. And if you don't know there's a problem and if you're not measuring it, then there's no way to improve on it, right? Because you can't, you don't know what you're measuring against. But I'm curious of the three buckets that you guys measured. So you measured data search and discovery was bucket number one, data literacy, you know, what do you do once you find it and then data governance in terms of managing. It feels like that the search and discovery, which is, it sounds like what you're primarily focused on is the biggest gap because you can't get to those other two buckets unless you can find and understand what you're looking for. So that, is that, is that jive or is that really the problem, is it more of the manipulation of the data once you get it? Yeah, I mean we focus really, we focus on all three. And I think, you know, certainly it's the case that it's a virtuous cycle. So if you think about kind of search and discovery of data, if you have very little context, then it's really hard to guide people to the right bit of information. But if I know, for example, that a certain data is used by a certain team and then a new member of that team comes on board, then I can go ahead and serve them with exactly that bit of data because I know that the human relationships are quite tight in the context graph on the back end. And so that comes from basically building more context over time. Now that context can come from a stewardship process, you know, implemented by a data governance framework. It can come from, you know, building better data literacy through having more analytics, but however that context is built and revealed, you know, there tends to be a virtuous cycle, which is you get more people searching for data. Then once they've searched for the data, you know how to necessarily build up the right context. And that's definitely done through data governance and data stewardship. And then once that happens, you're building literacy in the organization. So people then know what data to search for. So that tends to be a cycle. Now, often people don't recognize that cycle. And so they focus on one thing, thinking that you can do one to the exclusion of the others, but of course that's not the case. You have to do all three. Right, and I would presume you're using some good machine, machine learning and artificial intelligence in that process to continue to improve it over time as you get more data, the metadata around the data in terms of the usage. And I think again, I saw in another interview you did talking about, you know, where should people invest? What is the good data? What's the crap data? What's the stuff we shouldn't use because nobody ever uses it? Or what's the stuff maybe we need to look and decide whether we want to keep it or not versus, you know, the stuff that's guiding a lot of decisions with Bob, Mary and Joe, that seems to be a good investment. So, you know, it's a great application of applied AI and machine learning to a very specific process to again, get you in this virtual cycle, that sounds awesome. Yeah, no, it is. And it's really helpful to, I mean, it's really helpful to think about this. I mean, one of the big problems with data is that it's so abstract, but it's really helpful to think about it in just terms of use cases. Like if I'm using a customer data set and I want to join that with a transaction data set, just knowing which other transaction data sets people joined with that customer data set can be super helpful if I'm an analyst coming in to try to answer a question or ask a question. And so context can come in different ways, you know, just in the same way that Amazon does people who bought this product also bought this product, you can have all of the same analogies exist. People who use this product also use that product. And so being able to generate all that intelligence from the backend to serve up a simple seeming experience on the front end is the fun part of the problem. I'm just curious, because there's so many pieces of this thing going on. What's kind of the aha moment when you're in with a new customer and you know, you finish the install and you know, you've done all the crawling and you know where all the data sets are and you've got some baseline information about who's using what. I mean, what is kind of the, oh my goodness, when they see this thing suddenly delivering results that they've never had at their fingertips before? Yeah, you know, it's so funny because you can show Alation as a demo and you can show it to people with data sets that are fake. And so we have this like medical provider data set that you know, we've got in there and we've got a whole bunch of other data sets that are in there and people look at it and you know, interestingly enough, a lot of the time they're like, oh yeah, I can kind of see it work and I can kind of like understand that. And then you turn it on against their own data. The data they've been using every single day and literally their faces change. They look at the data and they say, oh my God, like this is a data set that Steven uses. I didn't even know that Steven thought that this data existed. And oh my God, like people are using this data in this particular way. They shouldn't be using that data at all. Like I thought I deprecated that data set two years ago. And so people, you know, have all of these interesting insights and it's interesting how much more real it gets when you turn it on against the company systems themselves. And so that's been a really fun thing that I just seen over and over again over the course of multiple years where people just turn on the product and all of a sudden it just changes their view of how they've been doing it all along. And that's been really fun and exciting. Great, yeah, because it means something to them, right? It's not numbers on a page. It's actually, it's people, it's customers, it's relationships, it's a lot of things. That's a great story. And I'm curious too, in that process, is it more often that they just didn't know that there were these other buckets of reports and other buckets of data, or was it more that they just didn't have access to it or if they did, they didn't really know how to manipulate it or to integrate it into their own workflow? Yeah, it's kind of funny and it's somewhat role dependent, but it's kind of all of the above. So if you think about it, if you're a data management professional, often you kind of know what data sources might exist in the enterprise, but you don't necessarily know how people are using the data. And so you look at the data and you're like, oh my God, I can't believe this team is using this data for this particular purpose. They shouldn't be doing that. They should be using this other data set. I deprecated that data set like two years ago. And then sometimes if you're a data scientist, you find, oh my gosh, there's this new database that I otherwise didn't realize existed. And so now I can use that data and I can process that for building some new machine learning algorithms. In one case, we found a customer where they had the same data set procured five different times. So it was a data set that caused multiple hundreds of thousands of dollars. They were spending $2 million overall on a data set where they could have been spending literally one fifth of that amount. And then you had sort of another case, finally, where you're basically just looking at it and saying, hey, I remember that data set. I knew I have that data set, but I just don't remember exactly where it was. Where did I put that report? And so it's exactly the same way that you would use Google. Sometimes you use it for knowledge discovery, but sometimes you also use it for just remembering the thing you forgot. But the thing, like I remember when people were trying to put Google search in at companies just to find records, not necessarily to support data efforts. And the knock was always, you didn't have enough traffic to drive the algorithm to really have effective search, say across a large enterprise that has a lot of records, but not necessarily a lot of activity. So that's a similar type of problem that you must have. So is it really extracting that extra context of other people's usage that helps you get around kind of that? You just don't have a big numbers. Yeah. I mean, that kind of is fundamentally the special sauce. I mean, I think a lot of data management has been this sort of manual brute force effort where I get a whole bunch of consultants or a whole bunch of people in the room and we do this big documentation session and all of a sudden we hope that we've kind of painted the Golden Gate bridges at work, but knowing that three to six months later, you're going to have to go back and repaint the Golden Gate Bridge all over again, if not immediately, depending on the size and scale of your company. The one thing that Google did to sort of crawl the web was to really understand, oh, if a certain webpage was linked to super often, then that webpage is probably a really useful webpage. And when we crawl the logs, we basically do the exact same thing. And that's really informed getting a really, really specific day one view of your data without having to have a whole bunch of manual effort. And that's been really just magical. I mean, it's been, it's allowed people to really see their data very quickly in new different ways. And I think a big part of this is just friction reduction, right? We'd all love to have an organized data world. We'd love to organize all the information in a company, but for anybody who has an email inbox, organizing your own inbox, let alone organizing every database in your company just seems like a sysophysian effort. And so being able to focus people on what's the most important thing has been the most important thing. And that's kind of why we've been so successful. I love it. And I love just kind of the human factors kind of overlay that you've done to add the metadata with the knowledge of who is accessing these things and how are they accessing it? And the other thing I think is so important Satya, and as we talk about innovation all the time, everybody wants more innovation and they've got DevOps so they can get software out faster, et cetera, et cetera. But I fundamentally believe in my heart of heart that it's much more foundational than that. That if you just give more people access to more information and then the ability to manipulate and glean knowledge out of that information and then actually take action and have the power and the authority to take action and you have that across everyone in the company or an increasing number of people in the company, now suddenly you're leveraging all those brains. You're leveraging all that insight, you're leveraging all that kind of first line experience to drive kind of a DevOps type of innovation with each individual person as opposed to kind of classic waterfall with the Chief Innovation Officer doing PowerPoints in his office on his own time and then coming down from the mountain and handing it out to everybody to go build. So it's a really kind of paradox that by adding more human factors to the data, you're actually making it so much more usable and so much more accessible and ultimately more valuable. Yeah, it's funny. There's this new term of art called data intelligence. And it's interesting because there's lots of people who are trying to define it and there's this idea and I think IDC has got a definition and you can go look it up. But if you think about the core word of intelligence, it basically devolves down to the ability to acquire information or skills, right? And so if you then apply that to companies and data, data intelligence then stands to reason it's sort of the ability for an organization to acquire information or skills leveraging their data. And that's not just for the company but it's for every individual inside of that company. And we talk a lot about how much change is going on in the world with COVID and with wildfires here in California. And then obviously with the elections and then with new regulations and with preferences because now that COVID's happened, everybody's at home. So what products and what services do you have to deliver to them? And all of this change is basically what every company has to keep up with is truth is to survive. If capitalism is creative destruction, the world's getting destroyed like unfortunately more often than we'd like it to be. And so then you're sitting there going, oh my God, how do I deal with all of this? And it used to be the case that you could just build a company off of being really good at one thing. Like you could just be the best like logistics delivery company. But that was great yesterday when you were delivering to restaurants but since there are no restaurants in business, you just had to change your entire business model and be really good at delivering to homes. And how do you go do that? Well, the only way to really go do that is to be really, really intelligent throughout your entire company. And that's a function of data. That's a function of your ability to adapt to a world around you. And that's not just some CEO because literally by the time it gets the CEO it's probably too late. Innovation's got to be occurring on the ground floor and people got to repackage things really quickly. I love it, I love it. And I love the other human factor that we talked about earlier is just people are curious, right? So if you can make it easy for them to fulfill their curiosity, they're going to naturally seek out the information and use it versus if you make it painful like a no-fund lesson, then people's eyes roll and they don't pay attention. So I think that it's such an insightful way to address the problem and really the opportunity. And the other piece I think that's so different when you're going down the card catalog analogy earlier, is there was a day when all the information was in that library. And if you went to the UCLA psych library, every single reference that you could ever find is in that library. I know I've been there, it was awesome. But that's not the way anymore, right? You can't have all the information and it's pulling your own information along with public information and as much information as you can where you start to build that competitive advantage. So I think it's a really great way to kind of frame this thing where information in and of itself is really not that valuable. It's about the context, the usability, the speed of the usability and that democratization is where you really start to get these force multipliers and using data as opposed to just talking about data. Yeah, and I think that that's the big insight, right? Like if you're a CEO and you're kind of looking at your chief data officer or chief data and analytics officer, the real question that you're trying to ask yourself is, how often do my people use data? How measurable is it? Like how much do people, what is the level at which people are making decisions, leveraging data? And that's something that you can talk about in a board room and you can talk about in a management meeting, but that's not where the question gets answered. The question gets really answered in the actual behaviors of individuals and the only way to answer that question if you're a chief data analytics officer or somebody who's responsible for data usage within the company is by measuring it and managing it and training it and making sure it's a part of every process and every decision by building a habit and building those habits are just super hard. And that's I think the thing that we've chosen to be sort of the best in the world at. And it's really hard. I mean, we're still learning about how to do it, but from our customers and taking that knowledge and kind of learning about it over time. Right, well, that's fantastic. And if it wasn't hard, it wouldn't be valuable. So those are always the best problems to solve. So Satya, really enjoyed the conversation. Congratulations to you and the team on the new release. I'm sure there's lots of sweat, blood and tears that went into that effort. So congrats on getting that out and really great to catch up. Look forward to our next catch up. You too, Jeff, it's been great to talk. Thank you so much. All right, take care. All right, he's Satya and I'm Jeff. You're watching theCUBE. We'll see you next time. Thanks for watching.