 From Chicago, it's the Cube, covering Veritas Vision Solution Day 2018. Brought to you by Veritas. Welcome back to Chicago everybody. This is the Cube, the leader in live tech coverage. We're here on the ground covering the Veritas Vision Solution Days in Chicago. We just a couple of weeks ago, we were in New York City at the iconic, iconic Tavern on the Green. We're here at the Palmer House Hotel. Jyoti Swarup is here, the Vice President of the Global Marketing for Veritas, great to see you again. Thanks, Dave, glad to be here. A few weeks ago, we saw you in New York. Since then, you've been around the globe, talking to customers. You just gave a great presentation to about 60, 70 customers here in Chicago. Obviously, a lot of your customers here, New York, one of the big NFL cities. So, what have you learned in the last couple of weeks? Well, a lot. It's been exciting, right? Since New York, I've been in Dubai, Milan, Rome. It's all over the place. Sounds exciting, but a lot of jet lag and travel. But a lot of exciting customers with interesting challenges that we can solve for. But I guess I would summarize it into three parts. Obviously, there are data protection challenges that we solve at Veritas and have done so for 20 years. There are a lot of storage challenges that we talked about and how they're moving to the cloud and how we can assist with that. And lastly, the interesting thing is the whole compliance and AI and ML related challenges. As to how do they look ahead while staying compliant with what they have already? There are some major trends forcing people to rethink their data protection strategies. Obviously, cloud is one, the whole security and data protection world's coming together. The edge, just the whole distributed data trend. Machine intelligence is another one. There are things that you can do with all that data. Machine plus data plus cloud really changes the game. You guys have some hard news in that area. Bring us up to date. What are you announcing? So we're announcing Veritas predictive insights. Really excited about this announcement because when I joined Veritas about 16 months ago, I felt like Veritas sits on top of all these exabytes of data. We protect the largest number of exabytes of data. So we have access to the metadata of that data. So my question to the engineering team was, what are we doing with that metadata? Are we going to use it, leverage it so our customers can benefit from it from all this usage data that we get from other customers? And the answer was, yes, we're working on something. Hold on, you're new, right? And now we have it. So at Veritas, yes, it takes 12 to 16 months to build something at scale, right? We have hundreds of engineers that worked on this. So what we've done now is, especially with our appliances portfolio, we're able to give our customers intuitive predictive and proactive maintenance and support of their systems. And what does that mean? It means firmware upgrades, patches, things like that. They don't have to be personalized, flying and engineering in, engineering in to do kind of things. They can be automated. Oracle recently at Oracle Open World announced this whole autonomous database. Why can't data protection be autonomous, right? So that's how we think, right? Make everything autonomous, make everything predictive and proactive, and that's what predictive insights is about. So let's unpack that a little bit. So what are the enablers that allow you to actually take this next step? Obviously, you've got the data. You've got a classification engine that allows you to put data in buckets, if you will. Explain what that is and why it's important. I'm glad you brought up the classification engine because that was at the heart of everything Veritas did for the last 20 years, right? We called it Vic Veritas Information Classifier, where we classified all of the data that came in on ingest, unlike other people, other customers and other vendors. We classified all the data that came in from net backup and we told our customers, here's PII numbers, here's sensitive information, here's structured data, here's unstructured data. We did this really well for a long time. Now we wanted to take that to the next level, right? We wanted to tell our customers, what's actually going on with your infrastructure? You've classified the data, you've put it in here, what can you do with it next? Where can you put it? Can you optimize it after the cloud? How much will you pay for it? Can you remove something off of tape? How much do you pay for that? Can you put something in a long-term retention on-prem? How much would that cost you? So we would not only want to give them information about the classification of that data, but how to monetize that data, how much money would it cost to store that data in different areas? So this is a case where if you go back to the, some of you might remember, 2006, the Federal Rules of Civil Procedure, mandated that you were able to recover and deliver to a court of law electronic records. Well data classification was a critical component there. This is one of those cases like, if you've got an older athlete like Tom Brady, maybe he's not as fast as he used to be, but he's got it all up here, he knows the place before he sees it. You guys have the experience around things like data classification, which are table stakes to allow you to do this, but it's still a challenge for many folks in the industry. It's a metadata problem, isn't it? Yep, it absolutely is. It is a metadata problem, and it's a metadata advantage for us at Veritas because we sit on top of the highest amount of metadata. So how do I take advantage of the Veritas predictive insights? Where does it live? So we've announced it will be out there, GAID, the beginning of the year, 2019. We're rolling it out with our appliances portfolio first, because we have more control over it because the appliances and the hardware have been integrated with our software. So we give our customers predictive insights on all of their appliances that they buy from Veritas and their systems. Going forward, we'll extend that to our software only, sales motions as well, as extending it to other software platforms and other hardware platforms from other vendors as well. So we're working on some integrations that I can't talk about today, but we want to essentially take predictive insights and move it beyond Veritas in the future. Okay, so talk a little bit more about how it works. You're using machine learning technology. You're building models and training the data for different customers. How does it all actually come to fruition? Sure, so the first thing is, we're generating what we're calling SRS, our system reliability score. So our engine processes all of this information that comes from our customers' data, the usage data, and maps it to the hundreds of other customers, thousands of other customers' usage data that we have, to find patterns. So for example, if a disk hasn't had a firmware upgraded and hasn't done so for months, we can predictively let the customer know this disk is going to fail if you didn't upgrade this, but that's not enough. We actually allow them to click a button and upgrade the firmware right there to that disk so it's done, right? So it's not only letting customers know that here's something that's going to go wrong, but here's how to fix it as well. That's just one example of what we can do. Well, that's key. It's like the old days, you have a pager and you get an alert, and then you got to go do something. You're saying you're actually building automation into the process. Right, it's like chatbots. You respond to the chatbot right there and it does the action for you. You don't actually have to go somewhere and figure it out. So you've got this SRS score. So what happens when you cross that threshold? It tells the system, okay, take some remedial action or does it allow the customer to sort of make that choice? What's yours? Sure, so the SRS score is like your credit score, right? There's a lot of complexity underneath that score. So at the highest level, we tell the customer, if your score is about a certain point, your systems are healthy. They're running well. If they go below a certain point, right, let's say a 700 score and a credit score, you got to go watch out, why did it go below? And we'll give them the 10 or 20 reasons why the score went down, whether it's a firmware thing or a support issue or a hard drive issue. We tell them exactly what's about to go wrong so they can go fix it before it actually goes wrong. What do you, actually, before we go there, just some examples, some use cases that you expect in the field that you can talk to customers about. Give us some more. Sure, so we talked to a lot of companies with massive data centers. So one of the things they say is with our appliances, simple things like temperature changes. I was in Dubai, look, the temperature there can be crazy, it goes over 100 degrees Fahrenheit. So they say simple things like temperature changes can have massive effect on your hard drives and how that works. If my AI and ML algorithms or my software can practically tell me the temperature's going up, this is what's going to happen, increase the cooling, do something different, move the data somewhere, back it up, that's great for the customer. Can I take action just based on a simple thing like temperature? There was another interesting customer here in New York, actually, that came to me and said, we had this problem like every so many weeks, their disks would fail. And they thought it was due to temperature because it was in the summer, it wasn't. And after a lot of research turns out, it was the fire alarms that were going off. So the fire alarm, the fire alarm testing that was going on was actually causing disks to fail. Because of the vibration? For the vibration and the decibel level, it was interesting, right? And now our AI and ML knows that, so it's recorded, we know it and we'll be better off going forward, right? We'll tell other customers now that have data centers with massive, loud, high decibel fire alarms that this could be a potential issue, I'm not saying that is the issue, but this could be a potential issue that they would have never thought of otherwise. So what do you expect the business impact to be when you talk to customers about this capability under non-disclosure, et cetera? How are they seeing this impacting their business? So it's three things, right? Proactive support and maintenance. That's really important. Their customers are tired of talking to large vendors where the support connections are horrible, right? They have to go in and raise a ticket and do certain things and then they will ship a guy over to their site who'll come and fix it. That's just too long, slow and reactive. We want to make this proactive and autonomous. That's number one. Number two is total cost of ownership, right? So when customers are able to predict these failures, they don't have to have a certain money set aside for solving problems when they occur. They're like, I know this problem's coming up. I need to budget for it. So their TCO models get better and more predictable, right? And last but definitely not the least, when we extend this to beyond Veritas, they will be able to do more with their data. Again, what is that more? We don't know yet today, but when we are able to extend this to beyond Veritas, customers will be able to do a lot more with their data centers. So a couple of things this plays into, obviously digital transformation is all about being on all the time. You don't want to have, you don't want planned downtime or unplanned downtime. This allows you to at least plan more effectively and potentially eliminate any downtime so your data is always accessible. And it's also cloud-like in that you're automating a lot of the either recovery from failures or you're pushing a button and saying, okay, remediate this, packs that so you don't have the failure. So that's a sort of cloud-like approach. So you said it's available first part of 19 and it's available, is it in appliances or how do I get this? So we'll be rolling it out in appliances first, all the Veritas appliances and then we'll extend it to software only as well as beyond Veritas going forward. Awesome, Jyoti, thanks very much for taking us through the new capability. AI brought to data protection and anticipating problems before they occur, remediating them in an autonomous way. Appreciate your time, thanks for coming back on. All right, keep it right there, everybody, we'll be back with our next guest. Right after this short break, you're watching theCUBE from Chicago, Veritas Vision Solution Days, we'll be right back.