(tranquil music) >> Alation was an early pioneer in solving some of the most challenging problems in so-called big data. Founded early last decade, the company's metadata management and data catalog have always been considered leading examples of modern tooling by customers and analysts alike. Governance is one area that customers identified as a requirement to extend their use of Alation's platform. And it became an opportunity for the company to expand its scope and total available market. Alation is doing just that today, announcing new data governance capabilities, and partner integrations that align with the market's direction of supporting federated governance. In other words, a centralized view of policy to accommodate distributed data in this world of an ever expanding data cloud, which we talk about all the time in theCUBE. And with me to discuss these trends and this announcement is Satyen Sangani, who's the CEO and co-founder of Alation. Satyen, welcome back to the CUBE. Good to see you. >> Thank you Dave, It's great to be back. >> Okay, so you heard my open, please tell us about the patterns that you were seeing in the market and what you were hearing from customers that led you in this direction and then we'll get into the announcement. >> Yeah, so I think there are really two patterns, right? I mean, when we started building this notion of a data catalog, as you said a decade ago, there was this idea that metadata management broadly classified was something that belonged in IT, lived in IT and was essentially managed by IT, right? I always liken it to kind of an inventory management system within a warehouse relative to Amazon.com, which has obviously broadly published for the business. And so, with the idea of bringing all of this data directly to the business and allowing people arbitrarily, depending on their role to use the data. You know, you saw one trend, which was just this massive, shift in how much data was available at any given time. I think the other thing that happened was that at the same time, data governance went through a real transitionary phase where there was a lot of demand often spurred by regulations. Whether that's GDPR, CCPA or more recently than that, certainly the Basel accord. And if you think about all of those regulations, people had to get something in a place. Now what we ended up finding out was when we were selling in to add accounts, people would say, well guess what? I've got this data governance thing going on, but nobody's really using it. I built this business glossary, it's been three years. Nothing's been really very effective. And we were never able to get the value and we need to get value because there are so many more people now accessing and using and leveraging the data. And so with that, we started really considering whether or not we needed to build a formal capability in the market. And that's what we're today that we're doing. >> I liked the way you framed that. And if you think back, we were there as you were in the early big day-to-days. And all the talk was about volume, variety and velocity. And those are sort of IT concepts. How do you deal with all these technical challenges? And then the fourth V which you just mentioned was value. And that's where the line of business really comes in. So let's get into the news. What are you announcing today? >> So we're announcing a new application on top of Alation's Catalog platform, which is an Alations data governance application. That application will be released with our 2021.3 release on September 14th. And what's exciting about that is that we are going to now allow customers to discreetly and elegantly and quickly consume a new application to get data governance regimes off the ground and initiatives off the ground, much more quickly than they've ever been able to do. This app is really all about time to value. It's about allowing customers to be able to consume what they need when they need it in order to be able to get successful governance initiatives going. And so that's what we're trying to deliver. >> So maybe you could talk a little bit about how you think about data governance and specifically your data governance approach. And maybe what's different about Alation's solution. >> Yeah, I think there's a couple of things that are different. I think the first thing that's most critically different is that we move beyond this notion of sort of policy declaration into the world of policy application and policy enforcement, right? I think a lot of data governance regimes basically stand up and say, look you know, it's all about people and then process and then technology. And what we need to do is declare who all the governors are and who all the stewards are. And then we're going to get all our policies in the same place and then the business will follow them. And the reality is people don't change their workflows to go off and arbitrarily follow some data governance policy that they don't know exists, or they don't want to actually have to follow up. And so really what you've got to do is make sure that the policy and the knowledge exists as in where the data exists. And that's why it's so critical to build governance into the catalog. And so what we're doing here is we're basically saying, look, you could declare policies with a new policy center inside of Alation. Those policies will get automatically created in some cases by integrating with technologies like Snowflake. But beyond that, what we're also doing is we're saying, look, we're going to move into the world of taking those policies and applying them to the data on an automated basis using ML and AI and basically saying that now it doesn't have to be some massive boil the ocean three-year regime to get very little value in a very limited business loss rate. Rather all of your data sets, all of your terms can be put into a single place on an automated basis. That's constantly being used by people and constantly being updated by the new systems that are coming online. And that's what's exciting about it. >> So I just want to follow up on that. So if I'm hearing you correctly, it's the humans are in the loop, but it's not the only source of policy, right? The machines are assisting. And in some cases managing end-to-end that policy. Is that right? >> You've got it. I think the the biggest challenge with data governance today is that it basically relies a little bit like the Golden Gate Bridge. You know, you start painting it and by the time you're done painting it, you've got to go back and start painting it again, because it relies upon people. And there's just too much change in the weather and there's too much traffic and there's just too much going on in the world of data. And frankly in today's world, that's not even an apt analogy because often what happens is midway through. You've got to restart painting from the very beginning because everything's changed. And so there's so much change in the IT landscape that the traditional way of doing data governance just doesn't work. >> Got it, so in winning through the press release, three things kind of stood out. I wonder if we could unpack them, there were multi-cloud, governance and security. And then of course the AI or what I like to call machine intelligence in there. And what you call the people centric approach. So I wonder if we could dig in into these and help us understand how they fit together. So thinking about multi-cloud governance, how do you think about that? Why is that so challenging and how are you solving that problem? >> Yeah, well every cloud technology provider has its own set of capabilities and platforms. And often those slight differences are causing differences in how those technologies are adopted. And so some teams optimize for certain capabilities and certain infrastructure over others. And that's true even within businesses. And of course, IT teams are also trying to diversify their IT portfolios. And that's another reason to go multi-cloud. So being able to have a governance capability that spans, certainly all of the good grade called megascalers, but also these new, huge emerging platforms like Snowflake of course and others. Those are really critical capabilities that are important for our customers to be able to get a handle on top of. And so this idea of being cloud agnostic and being able to sort of have a single control plane for all of your policies, for all of your data sets, that's a critical must have in a governance regime today. So that's point number one. >> Okay and then the machine learning piece, the AI, you're obviously injecting that into the application, but maybe tell us what that means both maybe technically and from a business stand point. >> Yeah, so this idea of a data policy, right? Can be sometimes by operational teams, but basically it's a set of rules around how one should and should not be able to use data, right? And so those are great rules. It could be that people who are in one country shouldn't be able to access the data of another country, very simple role, right? But how do you actually enforce that? Like you can declare it, but if there is a end point on a server that allows you to access the data, the policy is effectively moot. And so what you got to go do is make sure that at the point of leverage or at the point of usage, people know what the policy happens to be. And that's where AI come in. You can say let's document all the data sets that happened to be domiciled in Korea or in China. And therefore make sure that those are arbitrarily segregated so that when people want to use that as datasets, they know that the policy exists and they know that it's been applied to that particular dataset. That's somewhere where AI and ML can be super valuable rather than a human being trying to document thousands of databases or tens of thousands of data sets, which is really kind of a (mumbles) exercise. And so, that application of automation is really critical and being able to do governance at the scale that most enterprises have to do it. >> You got it 'cause humans just can't do that at scale. Now what do you mean by people-centric approach? Can you explain that? >> Yeah, often what I find with governance is that there's this notion of kind of there's this heavy notion of how one should deal with the data, right? So often what I find is that there are certain folks who think, oh well, we're going to declare the rules and people are just going to follow them. And if you've ever been well, a parent or in some cases seeing government operate, you realize that that actually isn't how things work. And involve them in how things are run. And if you do that, right? You're going to get a lot more success in how you apply rules and procedures because people will understand that and people know why they exist. And so what we do within this governance regime is we basically say, look, we want to make sure that the people who are using the data, leveraging the data are also the people who are stewarding the data. There shouldn't be a separate role of data steward that is arbitrarily defined off, just because you've been assigned to a job that you never wanted to do. Rather it should be a part of your day job. And it should be something that you do because you really want to do it. And it's a part of your workflow. And so this idea of being people centric is all about how do you engage the analyst, the product managers, the sales operation managers, to document those sales data sets and those product data sets. So that in fact, those people can be the ones who are answering the questions, not somebody off to the side who knows nothing about the data. >> Yeah, I think you've talked in previous CUBE interviews about context and that really fits to this discussion. So these capabilities are part of an application, which is what? it's a module onto your existing platform. And it's sort of it's a single platform, right? I mean, we're not bespoke products. Maybe you can talk about that. >> Yeah, that's exactly right. I mean, it's funny because we've evolved and built a relation with a lot of capability. I mean, interestingly we're launching this data governance application but I would say 60% of our almost 300 customers would say they do a form or a significant part of data governance, leveraging relations. So it's not like we're new to this market. We've been selling in this market for years. What's different though, is that we've talked a lot about the catalog as a platform over the last year. And we think that that's a really important concept because what is a platform? It's a capability that has multiple applications built on top of it, definitionally. And it's also a capability where third party developers can leverage APIs and SDKs to build applications. And thirdly, it has all of the requisite capabilities and content. So that those application developers, whether it's first party from Alation or third party can really build those applications efficiently, elegantly and economically well. And the catalog is a natural platform because it contains all of the knowledge of the datasets. And it has all of the people who might be leveraging data in some fundamental way. And so this idea of building this data governance module allows a very specialized audience of people in governance to be able to leverage the full capabilities of the platform, to be able to do their work faster, easier, much more simply and easily than they ever could have. And that's why we're so excited about this launch, because we think it's one example of many applications, whether it's ourselves building it or third parties that could be done so much more elegantly than it previously could have been. Because we have so much knowledge of the data and so much knowledge of how the company operates. >> Irrespective of the underlying cloud platform is what I heard before. >> irrespective of the underlying cloud platform, because the data as you know, lives everywhere. It's going to live in AWS, it's going to live in Snowflake. It's going to live on-premise inside of an Oracle database. That's not going to be changed. It's going to live in Teradata. It's going to live all over the place. And as a consequence of that, we've got to be able to connect to everything and we've got to be able to know everything. >> Okay, so that leads me to another big part of the announcement, which is the partnership and integration with Snowflake. Talk about how that came about. I mean, why snowflake? How should customers think about the future of data management. In the context of this relationship, obviously Snowflake talks about the data cloud. I want to understand that better and where you fit. >> Yeah, so interestingly, this partnership like most great partnerships was born in the field. We at the late part of last year had observed with Snowflake that we were in scores of their biggest accounts. And we found that when you found a really, really large Snowflake engagement, often you were going to be complementing that with a reasonable engagement with Alation. And so seeing that pattern as we were going out and raising our funding route at the beginning of this year, we basically found that Snowflake obviously with their Snowflake Ventures Investment arm realized how strategic having a great answer in the governance market happened to be. Now there are other use cases that we do with Snowflake. We can certainly get into those. But what we realized was that if you had a huge scale, Snowflake engagement, governance was a rate limiter to customers' ability to grow faster. And therefore also Snowflake's ability to grow faster within that account. And so we worked with them to not only develop a partnership but much more critically a roadmap that was really robust. And so we're now starting to deliver on that roadmap and are super excited to share a lot of those capabilities in this release. And so that means that we're automatically ingesting policies and controls from Snowflake into Alation, giving full transparency into both setting and also modifying and understanding those policies for anybody. And so that gives you another control plane through which to be able to manage all of the data inside of your enterprise, irrespective of how many instances of Snowflake you have and irrespective of how many controls you have available to you. >> And again, on which cloud runs on. So I want to follow up with that really interesting because Snowflake's promise of the data cloud, is it essentially abstracts the underlying complexity of the cloud. And I'm trying to understand, okay, how much of this is vision, how much is is real? And it's fine to have a Northstar, but sometimes you get lost in the marketing. And then the other part of the promise, and of course, big value proposition is data sharing. I mean, I think they've nailed that use case, but the challenge when you start sharing data is federated governance. And as well, I think you mentioned Oracle, Teradata that stuff's not all in the cloud, a lot of that stuff on-prem and you guys can deal with that as well. So help us sort of to those circles, if you can. Where do you fit into that equation? >> I think, so look, Snowflake is a magical technology and in the sense that if you look at the data, I mean, it reveals a very, very clear story of the ability to adopt Snowflake very quickly. So any data team with an organization can get up and running with the Snowflake instance with extraordinary speed and capability. Now that means that you could have scores, hundreds of instances of Snowflake within a single institution. And to the extent that you want to be able to govern that data to your point, you've got to have a single control plane through which you can manage all of those various instances. Whether they're combined or merged or completely federated and distinct from each other. Now, the other problem that comes up on governance is also discoverability. If you have all these instances, how do you know what the right hand is doing if the left hand is working independently of it? You need some way to be able to coordinate that effort. And so that idea of discoverability and governance is really the value proposition that Alation brings to the table. And the idea there is that people can then can get up and running much more quickly because, hey, what if I want to spin up a Snowflake instance, but there's somebody else, two teams over those already solved the problem or has the data that I need? Well, then maybe I don't even need to do that anymore. Or maybe I can build on top of that work to be able to get to even better outcome even faster. And so that's the sort of kind of one plus one equals three equation that we're trying to build with them. >> So that makes sense and that leads me to one of my favorite topics with the notion is this burgeoning movement around the concept of a data mesh in it. In other words, the notion that increasingly organizations are going to push to decentralize their data architectures and at the same time support a centralized policy. What do you think about this trend? How do you see Alation fitting in? >> Yeah, maybe in a different CUBE conversation. We can talk a little bit about my sort of stylized history of data, but I've got this basic theory that like everybody started out what sort of this idea of a single source of truth. That was a great term back in the 90s where people were like, look, we just need to build a single source of truth and we can take all of our data and physically land it up in a single place. And when we do that, it's going to all be clean, available and perfect. And we'll get back to the garden of Eden, right? And I think that idea has always been sort of this elusive thing that nobody's ever been able to really accomplish, right? Because in any data environment, what you're going to find is that if people use data, they create more data, right? And so in that world, you know, like that notion of centralization is always going to be fighting this idea of data sprawl. And so this concept of data mesh I think is, you know, there's formal technical definitions. But I'll stick with maybe a very informal one, which is the one that you offered. Which is just sort of this decentralized mode of architecture. You can't have decentralization if nobody knows how to access those different data points, 'cause otherwise they'll just have copies and sprawl and rework. And so you need a catalog and you need centralized policies so that people know what's available to them. And people have some way of being able to get conformed data. Like if you've got data spread out all over the place, how do you know which is the right master? How do you know what's the right customer record? How do you know what's your right chart of accounts? You've got to have services that exist in order to be able to find that stuff and to be able to leverage them consistently. And so, to me the data mesh is really a continuation of this idea, which the catalog really enabled. Which is if you can build a single source of reference, not a single source of truth, but a single place where people can find and discover the data, then you can govern a single plane and you can build consistent architectural rules so that different services can exist in a decentralized way without having to sort of bear all the costs of centralization. And I think that's a super exciting trend 'cause it gives power back to people who want to use the data more quickly and efficiently. >> And I think as we were talking about before, it's not about just the IT technical aspects, hey, it works. It's about putting power in the hands of the lines of business. And a big part of the data mesh conversation is around building data products and putting context or putting data in the hands of the people who have the context. And so it seems to me that Alation, okay, so you could have a catalog that is of the line of businesses catalog, but then there's an Uber catalog that sort of rolls up. So you've got full visibility. It seems that you've fit perfectly into that data mesh. And whether it's a data hub, a data warehouse, data lake, I mean, you don't care. I mean, that's just another node that you can help manage. >> That's exactly right. I mean, it's funny because we all look at these market scapes where people see these vendor landscapes of 500 or 800 different data and AI and ML and data architecture vendors. And often I get asked, well, why doesn't somebody come along and like consolidate all this stuff? And the reality is that tools are a reflection of how people think. And when people have different problems and different sets of experiences, they're going to want to use the best tool in order to be able to solve their problem. And so the nice thing about having a mesh architecture is you can use whatever tool you want. You just have to expose your data in a consistent way. And if you have a catalog, you can be able to have different teams using different infrastructure, different tools, different fundamental methods of building the software. But as long as they're exposing it in a consistent way, it doesn't matter. You don't necessarily need to care how it's built. You just need to know that you've got good data available to you. And that's exactly what a catalog does. >> Well, at least your catalog. I think the data mesh, it should be tools that are agnostic. And I think there are certain tools that are, I think you guys started with that principle. Not every data catalog is going to enable that, but I think that is the trend Satyen. And I think you guys have always fit into that. It's just that I think you were ahead of the time. Hey, we'll give you the last word. Give us the closing thoughts and bring us home. >> Well, I mean that's exactly right. Like, not all the catalogs are created equal and certainly not all governance is created equal. And I think most people say these words and think that are simple to get into. And then it's a death by a thousand cuts. I was literally on the phone with a chief data officer yesterday of a major distributor. And they basically said, look, like we've got sprawl everywhere. We've got data everywhere. We've got it in every type of system. And so having that sophistication turned into something that's actually easy to use is a super hard problem. And it's the one that we're focused on every single day that we wake up and every single night when we go to sleep. And so, that's kind of what we do. And we're here to make governance easy, to make data discovery easy. Those are the things that we hold our hats on. And we're super excited to put this release out 'cause we think it's going to make customers so much more capable of building on top of the problems that they've already solved. And that's what we're here to do. >> Good stuff, Satyen. Thanks so much, congratulations on the announcement and great to see you again. >> You too, Dave. Great talking. >> All right, thanks for watching this CUBE conversation. This is Dave Vellante, we'll see you next time. (tranquil music)