 Welcome everyone, this is theCUBE. My name is Paul Gillan, I'm the Enterprise Editor of SiliconANGLE, and today we have a CUBE conversation interview with Ash Ashtosh, who is the CEO of Actifio, a company that you may not have heard of, but which has raised a lot of money and is a unicorn company based in Massachusetts in the storage copy data virtualization arena. Now, if you've been around the storage industry for any period of time, you've probably heard of Ash. He's a 25-year veteran of storage, the storage business. He is the founder of a number of companies, including AppIQ, Serrano Systems, the author of numerous storage standards, and was a Greylock Partners as a VC, and then came over to head up Actifio. Also a CUBE alumnus, he has been on the CUBE before, about five years ago, I believe it was Ash, welcome. Thank you, Paul. Thanks for joining us today. A lot has happened in those five years. You've done on a growth trajectory. Why don't you tell us about the status of the company right now, and we'll talk a little bit more about your technology. Yeah, that's great. So just as an introduction, we pioneered this whole notion of copy data virtualization, and today Actifio delivers data virtualization where there's instant access to data that is inherently protected and completely independent of infrastructure, basically delivering the same capability that led us to start Actifio, based on the fact that servers were virtualized, networks are getting virtualized, it was time to get data to be virtualized. And over 1,900 of the global enterprises today, in 37 countries use Actifio, some of the largest financial services organizations, some of the largest service providers, 60 of the largest service providers in the world that use data as the fundamental resource that they go back and sell. And pretty much any organization with a decent amount of data, when we talk about decent amount, being 40 terabytes or above, or lots of people who use data, where digital data is the foundation of the business. So that's what we've been spending quite a bit of time on. So there's been a fundamental change in the way people think about infrastructure, and we believe when sheet metal and blinking lights have been replaced by APIs, it is time for making data than even for structure. Well, you say to make data a virtualization, to make a data virtual in the same way that we made networks and storage and servers virtual. What does that mean, data virtualization? That's a great question, right? So if you look back at the origins of the company about eight years ago, as we started seeing more and more businesses becoming digital, the transformation of every business to be a digital enterprise, changed the fundamental nature of what's most valuable for an enterprise, data and applications became the most valuable asset. They became the lifeblood of how I run my business. Whether it is Uber, the largest taxi company with no taxis, whether it is Airbnb, the largest hotel company with no hotels, or Rosetta Stone, one of our customers who used to sell languages, language translation, language learning software on CDs, now completely does it online with no assets. For these businesses, what used to be a very simple process of applications create data and ops operations people manage data, turned into applications create data that is then used to go back and develop new applications because I need to respond to my users. That is then again used to go back and create new analytics, that is then fed back into the application. What used to be a simple ops became dev ops and now has become a dev ops litics, we call it, it's a combination of dev ops and analytics. And it completes a cycle. That whole resource is based on one fundamental concept. It doesn't depend on servers. It doesn't count on networks. It doesn't rely on the fact that there's a specific storage. What it relies on is my data. And that is one I need to be able to access instantly. The ability for an application to come back and use an API to get any data I want from a single system of record for the enterprise. That's what active your developers. So think of us as this fundamental system of record for the entire enterprise where I can go tap in just like I call Amazon or AWS for a resource whether it's compute network or storage, you make an API call to active you, you get SAP from 14 days ago that 24 of my developers can go back and develop against. You can go back as a retail organization instead of creating another entity for data warehousing call an API and look at last year's data and run analytics, run Hadoop on it, run all kinds of data warehousing applications on it without having to go back and create another pool of storage and infrastructure. So for the first time, you have a single place where data is the new infrastructure and you really don't care what underlying infrastructure. This sounds, pardon me, this sounds a lot like a database management system to me, what's the difference? I think this is, data is more than databases. Right now, we have users who not only manage structured and unstructured data as part of the life cycle. We have people putting IOT data, social media data, data from other external resources but more importantly, there's a bunch of stuff that the operations people used to do to protect the data. There are things I have to do to back it up to make sure it is business from a business resilience perspective it's available. Make sure it is mobile, there's mobility built in so that it's available where I need it to be. We have one of the largest financial institutions today that has over 3,000 databases now, going up to 5,000 by the end of the year. Several hundred developers in eight countries access this data from anywhere with no concern for what infrastructure it's running on. There is no more dependency on operations person to give me the data I need to do what I need to do whether it's developing, fixing a bug or running analytics, very, very different from the way you used to think about it before. So database or even data warehousing it was an old model where you can- Data lake. Yeah, those were all dumping data in there but that's a perfect example of another copy I needed to make specifically for one operation versus I already have the data in the enterprise. We virtualize the whole thing in one single system of record that manages all aspects of the life cycle. And how are you protecting that data from corruption on the back end? I mean, we were talking about live data or operational transactional data operations people get pretty nervous about granting access for purposes that aren't related to the business transaction. That is a great question which is why we set out seven years ago and pioneered this whole motion called copy data. The reality is in the enterprise there are two kinds of data. One is production data, stuff that business applications are running on and you cannot impact this because this is what my transaction, this is what my, when I swipe my card this is where activity is going on. And then there's an entire business. In fact, a very valuable part of the business that uses copies of this data for development, for analytics, for protection, for disaster recovery. And we virtualize this entire part of it without ever touching anything on the production side. So we never go back and touch the production part. We make one copy, what we call the golden copy manage its entire life cycle through a distributor object file system that is independent of infrastructure that takes care of mobility that's required. And that allows for a very different way of thinking about what a storage system is. A lot of, I can get into the details of how this is done. But this is about moving the conversation to application objects that I need as opposed to storage learns or volumes. Because really at the end of the day I care about a file. I care about my SAP object. I care about six months of retail data then which learn and which volume it came from. And it was important to completely abstract this notion out. So what's the payoff here? Is it agility? Is it a faster application development? Is it savings on data storage? Absolutely, yeah. So began, so company's seven years old. First few years it was all about driving down the footprint of storage. So two dimensions, right? Footprint of the ability to bring in all this disparate copies down to one single golden copy that reduce the footprint. But second it was also very similar to the VMware part. It was about commoditizing storage. I didn't need all the fancy snapshot and replication that a storage array gave me. I just needed a reliable storage box that I could run my data management software anywhere on. So cost was a big part. And that was a big driving force. And we have organizations where we walked into in a five year period, we have taken out 292 million British pounds of hardware. Gone, gone because there was so much of redundancy all over the place, which was ridiculous. There were 8,000 developers accessing individual copies of data all over the place. There were silos of backup everywhere. This DR spread out all over the place. And you consolidated all that into a single system of record. That's number one. But even more than that, in the last two years, 43% of our business comes from agility. It's about speed of access to data so that I can develop faster, so that I can develop my applications faster. I can bring up my business applications faster. About two years ago, along with a service provider, we delivered the world's fastest business resiliency solution. You can bring up your entire data center in 20 minutes. Took you eight weeks. The ability to bring up that either for availability or for applications or analytics. Speed became big part. And as a result, if you think about consolidation, simplification, the third big part was driving down the complexity of this entire management process. Managing data lifecycle, as we know, the backup administrator was a hazing job. When you guy came in, they would make him a backup guy. And that was one job. But now for the first time, you have a chief data officer who can literally control per application the SLA that defines the entire lifecycle with one place, collectifio, and not have to deal with every piece of infrastructure anymore. So three parts, cost, speed, and there's the simplicity. And it's ironic that until four years ago or two years ago, cost was a big part. But we believe now, fast is the new big. Speed has become the most important part. So do customers know that they have a problem or do they know that they have an opportunity? Agility, as you're talking about. I mean, what is the awareness of the cycle? Like how much do you have to evangelize these ideas? Yeah, I mean, I think it took us four years to let the world know. This is the burden of creating a category is that you have to go out there and evangelize. And fortunately, the problem was so obvious that it took us five minutes to tell people how much it costs you and why it is such a big issue. But four years later, now you have every storage company and half a dozen startups and every backup company getting into the coffee data business, one form or the other. And I believe it's now, when I see some of the largest customers of ours saying, hey, we're putting out an RFP for coffee data, I think we know we are right. People have realized that this is as big a problem as they've seen on the virtualization of the computer side. That they go to solve this bigger problem. And it's not because of cost. We went from zero to 43% in less than two years on the DevOps and Agility and Hybrid Cloud side. We had no business in that less than two years ago. But people are beginning to realize, speed, application development, analytics is even more important than saving money. Well, when you talk about speed, I mean, can you quantify that? What do you mean by how much faster? Absolutely, I mean, so we have some of the largest healthcare companies that have well over 100 terabyte of single instance of a database. Single instance would take them four to six weeks to restore these databases, if ever. First of all, they couldn't be backed up. But when they were backed up, it would take them weeks to restore, less than five minutes. You can get an entire virtual instance of a database to come up. Three terabyte, five terabyte, 50 terabyte, 100 terabyte databases. You're talking about orders of magnitude of speed. And it's very simple. We just took out the whole business of, if you think about what we've done, we took the paradigm of, I used to have data on disk, and it needs to go on to tape. And so I need to do the translation, put it in the network, bring it up here. And when I need it to restore, I need to bring it back. We just eliminated the process of translating or transforming the data formats. And worse, ever having to bring it back. We look like a storage system. We look like we have a fiber channel, I SCSI and NAS interface. The only difference is when you connect to us, you determine what data you see behind. So unlike a traditional storage system where the data I see is what I wrote, and the storage system is a dumb device whose only context it has is blocks, we have the complete context. You point to Actifio and looks like a storage node, a virtual one, and you say, I want to see the way Oracle look like, or I want to see the way my entire data center look like 10 days ago. We instantly recreate that entire stuff and make your code and code learn or volume, look like what it was, synthesize it. And that's the difference. So a time machine. It is absolutely a time machine. And I've been in the storage business for a long time. I built it as part of the team that built the very first storage systems, did the SCSI protocol, fiber channel protocol. And one of the things we did very well was based on a model that a compute and storage were pretty tightly connected and there was very little routing in between, very little context required. All the context and routing was left up to the networking guys. And as you got to the point where your data was decoupled from infrastructure, it was going everywhere, you needed to bring in the notion of a context. You needed systems to know which applications they came from, what time they came from, what's the life cycle of this object, and needed to be managed completely seamlessly. So think about what we've done is to go fix something that we kind of made a mistake on 25 years ago and entire storage protocols and created a new one. Well, now with the emergence of Hadoop and hyperconvergence, we're seeing a move back to positioning the data, taking the data back close to the computing resource again. How does that affect what you do? How do you play with that? And so, and I think as most people will say, the biggest gravity in IT is data. It's hard to move. You can bring up a server, you can bring up a VM, you can even bring up an application in seconds because there is no gravity to that. To bring data to where you need your application to run takes an enormous amount of time. People talk about bringing up, spinning up applications in AWS. Well, you can spin up an application, you just can't get your data fast enough. What we've done is to take care of the mobility of data whenever you need it to be. We built into Actifio is the notion of the world's most efficient way of moving data around. But more importantly, you can, with all the capabilities around, you can spin up any application you want, right where the data is, right where Actifio is. In addition to some of the data mobility part, we have a ton of applications of orchestration. It's not just we move data around. We literally, in business resiliency, we were talking about bringing up a data center for something that's failed on our site, failover site. We not only bring the data around, because we know the context, we know it's Oracle, we know it's IP addresses. We orchestrated the entire process of bringing up the applications, the IP addresses, take care that they come up in the appropriate order that's supposed to come up with. The ability to spin up an application, the ability to instantiate a VM, that is cheap. That is instant. The question is where do you want to bring up? We have large development organizations that, when they log in, some other developers end up developing their applications in an Oracle Cloud environment. Not on premises anymore. They don't even know that's happening. Some are still on premises. And this ability to make underlying sheet metal and blinking lights completely invisible, because you have data to be a transparent layer, that, I think, is the ultimate. You're talking about storing and managing a huge amount of data, though. Absolutely. How do you, what do you do to provide the scalability that you need to do that? That's a great question. It's very, very easy to build what active you build at a desktop level. Apple time machine is a great example. Well, that hard. It's very, very hard. It's an engineering challenge to build what we have built at the enterprise scale. We have some of the largest organizations we deal with petabytes of data. And it's all about building every step of the way. From the time we capture data, we are very efficient in capturing data at the application level. We have no context of what storage systems are underneath. We're very, very efficient at storing data. We're very efficient at moving data. And finally, very efficient at bringing up every step of this four-step journey. You have to be super efficient at managing. The context has to be about data, not about storage. So often you see a lot more copy data companies coming out with storage systems saying, hey, I have a snapshot and therefore it's copy data. It's a glorified snapshot manager. Did you write your own file system? Absolutely. If you have to do this, what you have to do is a distributed object file system that has one single global namespace. Because I should be able to access my data anywhere. But the context of what I'm accessing has to be in an object that is taken care of the workflow and the mobility across any location. This is hard. That took us a lot of time to get it. It's hard even to explain. It is hard. Yeah, but the part that we make it very simple is go call an API or use the application and access your data anywhere. And as you, you will never lose the data unless you want to delete it. Now, you have raised $225 million. Last round was over two years ago. Your investors appear to be very patient. Where are you looking to take this business in terms of a liquidity event or what's the end game? So the end game is simple. End game is there's an opportunity to create a sustaining business that fundamentally changes the entire storage slash IT market. If you notice what's happening in the entire storage market, this whole digital transformation was supposed to create an explosion for storage industry. It was supposed to create so much opportunity because all that digital data needed to be stored somewhere. If you remember 10 years ago, what turns out is create an implosion of storage industry. Finally, EMC has been acquired. And we see that the old model of storage is going to be replaced by data. And we see an opportunity for us to go back and start. And we chose to start at the top, the toughest problem, which is go after the enterprises that have the most amount of data that are building the biggest of the businesses around data and start there. And so if you look at our customer base, it is a global 2,000 accounts. I would say more than 2,000 accounts now. But that's where the stickiest part of the business is. We could have chosen to go after the SMB part, which is a lot more churn. We have a total of four users in the last seven years that we have had a churn on, four. I imagine once they make the investment, it's very difficult for them to decouple it even if they wanted to. But it's also the value they're getting. It's the number of people who are running on Activeio platform. There are 8,000 developers in the organization that use Activeio. You cannot take it out. By the way, it's also doing backup. It's also doing disaster recovery. It's also running analytics for some people who are going out and using cloud analytics. Not even my own. I'm using a third party analytics in the cloud. This ability to create, that's why this is what I meant by calling data as a new infrastructure. Because I have people who want data in one place that provides a system of record you can call anywhere you want or use the tools we provide and do what's required to build your digital business. So we believe there's a phenomenal opportunity to create this. We created a new category. We believe there's a new opportunity to come back and create a sustaining company. And so we shifted pretty heavily about a little over a year ago to focus on becoming a profitable company. And we'll see when the users want us to go public or not go public. I think the one thing you will see in Activeio, just by the nature of the kind of customers we are dealing with and the fact that we're dealing with data for them. This is a laser-focused company on customer success. Very laser-focused. So the day the customer tells you have to be a public company, maybe we will. Probably won't hear that from your investors before your customers. It's true, but it's ironic. Some of the largest companies today are private. Dell EMC is a private company. Well, the markets have not been kind lately. So you're biding your time. Yeah, we have no urgency. And I've always said this for years now. There are three constituencies that really care about being public. One, number one, by far is if our users say he looked transparency of your financials is important for us, that'll be the number one criteria. Obviously, shareholders become the number two part. And third one is comparative nature. And we don't see much of a comparative headwind so far. Well, we're out of time. Thank you so much for joining us. Thank you, Paul. Ashatash, great to see a company, Massachusetts-based company doing well, killing it in a field that has great growth, obviously, left to it. So that's why I hope you'll join us again soon. Thank you, Paul. I look forward to it. This is theCUBE. I'm Paul Gellin. Thanks for joining us. Thank you.