 from our studios in the heart of Silicon Valley, Palo Alto, California, this is a CUBE Conversation. Hey, welcome back everybody. Jeff Frick here with theCUBE. We're in our Palo Alto studios for a CUBE Conversation. We've got a really great guest. He's been on many, many times. We're always excited to have him on through a bunch of different companies a lot of years and really a great perspective. So we're excited to have Guy Churchward, the CEO of Dayterra back in the Palo Alto studio. Guy, great to see you. Thank you, Jeff, appreciate it. Absolutely. So I think last time you were here, I was looking it up actually, it was November of 2018. You were kind of just getting started on your Dayterra adventure. Give us kind of the update. Yeah, I was going to say last time we had Markin, who's CEO and founder, co-founder of Dayterra, and I was edging in. So I was executive chairman at the time. And obviously I found the technology. I was looking for an organization that had some forward thinking on storage. And we started to get very close with a large strategic and actually we announced it on the go-to market I think in February with HP. And I thought that myself and Mark kind of sat down, did a pinky swear and said, okay, maybe it's time for me to step in, take the CEO role, just to make sure that we had that sort of marriage of innovation and then some of the operation stuff that I could bring inside the business. So you've been at this for a while. You've been in the industry for a long time. What was it that you saw that really wanted you to get deeper in with Dayterra? Because obviously I'm sure you have tons of opportunities coming your way to kind of move from the board seat into the CEO position. Yeah, yeah, yeah, a bad bet, maybe stupidity or being drunk, to be honest, it was, the first thing is I was looking for this technology that basically spanned forward. And I had this gut hunch that organizations were looking for data freedom. That's why I did the data analytics job. Before that I did security analytics. And we were looking at that when we were, back when we talked to things like EMC and Dell. And so from a pure technology standpoint, I wanted to be in that space. But in the last few months, because you know, jobs are all about learning and then adjusting and learning and adjusting and learning and adjusting. And what I saw is a great bunch of guys, good technology, but we were sort of flapping around and had an idea that we were an advanced data services platform, it was to do with multi cloud. And in essence, I kind of came to this fundamental kind of understanding because I've been on both sides, which is Deterra is a bunch of cloud people trying to solve storage needs for what the cloud needs, but they have the experience, they've walked that mile. You know, when people always say, you've got to learn by walking in their shoes. Right, right. And they've done that versus where I've been in the past, where we were array specialists pushing towards a future that we didn't quite understand. You know, and there is a fundamental philosophical, philosophical difference between the two. And weirdly, my analogy or my aha moment came with the Tesla piece. And I know that, you know, you've pinged me a few times on, you know, Twitter over this. Right. I'm not a Tesla bigot to the extent of, you know, and probably I am now, I should have a Tesla t-shirt on. But I always thought it was an electric car and all they've done is electrified a car. And there was, and you know, I've resisted it for years and been not exactly an advocate. But I ended up buying one because I just, I felt from a technology standpoint in a platform that they were the right thing. And once I started to really understand what they were about, I saw these severe differences. And you know, we've chatted a little bit about this. And again, it's part of the analogy of what's happening in the storage industry but what's happening in the industry in a global position. But if you compare contrast something like Tesla to maybe Volkswagen, and it might be a bad example, but, you know, Audi, their first tranche into electric vehicles was the Audi A3. And I could imagine that they were traditional car people pushing their car forward, saw it as a combustion engine. Well, if I changed that and put some sort of powertrain in place which is an equivalent of a, you know, a system to basically drive the wheels and then a bunch of batteries, job done, all good. Right, right. And I assume that Tesla was the same. But I had a weird experience which is once you get it into autopilot, you can actually set the navigation direction and then it'll indicate, it'll hint to you when to change lanes. And so, for instance, I'm driving to the office and I'm going along 880 and I want to go onto 101. It says, you know, you need to pull across, I hit the indicator, it'll change lanes and it'll do some of the stuff. And that's all well and good. But I was up going to a board meeting on 280 going off at the Rosewood, you know, with Sandhill. Sandhill. And I was listening to a book, one of these, you know, audiobooks and I wasn't really paying much attention. I'm in the outside lane, obviously hitting the speed limit, no more. But I wasn't paying attention and all of a sudden the car basically indicates for me, changes lanes, slows down, changes lane again and then takes a junction, slows down, comes up to a junction. And you start to realize that actually Tesla's not about electrified vehicles, it's actually about the telemetry and the analytics and then feeding that back into the system. And I always thought a Tesla might be collecting how fast a car's going when they break, you know, the usual things. And everybody has this conversation, it's always over at work. But if you sort of look at it and you say, no, maybe they collect everything and then maybe what they're doing is they're collecting, hitting the indicator stalk. So when I'm coming up to a junction and I indicate, how long do I stay indicating before I break and then I change lanes and then I basically slow down and I go into the junction. And then what they do is they take that live information, crowdsource it, pull it back into the system and then when they're absolutely bulletproof, that junction then is exactly as a human would normally do this, they then let the car take over. So the difference between the two junctions is one, they totally understood. The other one they're still learning from. And then you look at it and you go, done. So they're basically an edge telemetry at a micro level organization. You know, and that is a massive difference between what Tesla's doing and a lot of the other car manufacturers are doing that are catching up. Which is really why I believe that they're going to be ahead for a long time. It's really interesting. I was an electronics wholesale for 10 years before I went back to school, can't go into the tech industry. And so really distribution was king. From the manufacturer point of view, always they just like ship their products for ages, right? They use distribution to break bulk. They use distribution to educate the customer. They use distribution just to get the stuff out. But they never knew how people actually operate their products, whether that be a car, a washing machine, a cassette player, whatever. So what fascinates me about these connected devices is what a fundamentally different set of data now manufacturers have, people have in how people actually use the product. But even more importantly, as you said, they can take that data and make adjustments on the fly because so much of it's software now. And we talked again before we turned on some of your software upgrades that you've gotten in the Tesla over the last six months which were all driven by customers. But they had a platform in place that enabled them to update functionality and to basically repurpose hardware elements for a new function, which is so in sync with DevOps and kind of this DevOps culture and this continuous upgrade, this continuous innovation with actual data from real people operating the products that they ship into the market. Yeah, and I think once you've stepped back and that was really why I was keen to sit down and talk and it's not specifically around software defined storage which is the day-to-day piece. And our example is yes, I'm the Tesla because we can do all of the analytics and all of the telemetry versus a standard array. And if you scratch that away and you say, let's have a look at our whole lives, our macro lives. Another example was my wife and I, we've got friends of ours are always banging on about these sleep by number beds. And so we went past the store, wandered in and the sales rep got us lying on a bed and he was doing the pumping the bed up to a size and he says, well, you are 65 or you are 70 or 75. And I kind of got bored of that. And I went, yeah, yeah, okay, I'm that. And he goes, okay, your wife's a 50 and you're a 75. And I said, well, that's kind of daft. And he goes, well, here's, and he shows then a map and it shows a thermal image of me lying on the bed. I'm a side sleeper or a back sleeper. And then what they do is they feed the information so that comes back off their edge, which is now a bed. And then what they do is they then analyze it and continuously prove it to try and include my bed, sleeping patterns. So you look at it and you say, what they're not doing is just manufacturing a mattress and throwing it out. What they've done is they said, we're going to treat each individual that lies on the mattress differently and we're going to take feedback and we're going to make that experience even better. So that same thing, which is this asset telemetry, micro-asset telemetry happens to be on the edge, is identical to what they have. And then I look at it and I go, why don't I like the array systems? Well, because the majority of stuff is I'm a file system. My brain is inherently looking at the drive types underneath and saying, as long as that works fine, everything that sits inside that, I don't care, it'll do its thing. And that was built around the whole process and premise of an application has a single function. But now applications create data. That data has multiple functions. And as people start to use it in different ways, you need to feed that data and the way in which it's processed differently. And so it all has that intelligence. Houses in home automation. I'm a junkie on anything that has a plug on it. And I've now got to a point where I have light switches or light fittings that have multiple bulbs. And every bulb now is actually can be, has telemetry around it, which I can adjust it dynamically based on the environment. Right, right. You know, I wish it got wine. You know, we've got the wine fridges. That's my biggest beef right now is you've got a wine fridge. You can have dual, you know, you have dual climates, which means that you don't fan to one side of it and they overheat at the bottom. But it'll break the grapes down. Wouldn't it be really cool if the cork actually had some way of figuring out what it needs to be fed and then each of them could be individual. But our entire being, you know, if you think about it's not just technology or technology's driving it, but it's not the IT industry, but our entire lives are now driven around exactly what you just described, which is manufacturers dropping something out into the wild to the edge and then having enough telemetry to be able to enhance that experience and then provide over the air, you know, enhancements. Right. And the other thing I think is fascinating, as it's looking up, we interviewed Derek Curtin from the AutoTech Council and that's a group locally that just works with all the municipalities and car manufacturers, tech companies. But he made a really interesting comment because there's the individual adjustment to you to know that you want to get off it at PageMill or Sandhill and you've got a calendar on your point this is me to the Rosewood. But then the other thing is when you aggregate that now back up, you know, not that you're going to be sharing other people's data, but when you start to get usage patterns from a large population that you can again incorporate best practices into upgrades of the product and you used a really good example and this was right after the one pedestrian got killed by the test of the lady with the bike that ran across the front of the street and it had literally happened a week before I think the conference. So very hot topic in Autonomous Vehicle Conference and what he said, which is really important, you know, if I get in an automobile accident, you know, I'm going to learn something, the person I hit is probably going to learn something, the insurance adjuster is going to take some notes and we're going to learn this bad intersection, I made a mistake, whatever. But when an autonomous vehicle gets in a wreck, and it's connected, all that telemetry goes back up into the system to feed the system, to make improvements for the whole system. So every car learns, every time one car has a problem, every time one car gets into a sticky situation. So again, kind of this crowdsourced learning and optimization opportunity is fundamentally different than I'm just shipping stuff out and I don't know what's going to happen to it and maybe a couple of pieces come back. So I think people that are not into the direct connection are so missing out on, as you said, this whole different level of data, this whole different level of engagement, a whole different level of product improvement and roadmap. It's not a PRD, it's not an MRD, it's all about get it out there, you know, get feedback from the usage and make those improvements on the fly. And continuous improvements and micro-analytics. I mean, even, you know, we talked back when you were adjusting how you deliver content for the cube, you know, rather than a big blob, you really want to say, well, I need more value for that, my clients need more value for that. So you've almost done that micro-segmentation by taking the information and then meta-tagging every single word in every single interview, to enrich the customer's experience. You know, and it kind of then you map back and you say, we've got to the age now where the IT staff, the executives that we talked to over the other side of the table, they're us, they live in our lives, they've got the same kids as we've got, the same ages we've got, they do the same purchasing as we've got. They understand the same things and they get frustrated when things naturally don't work the way they should. Like I've got a home theater system and I've still got three remote controls. I can't get down, I've got a universal remote control but it won't work because the components don't think. So what's happened is we've got to a world where everything's kind of interconnected and everything kind of learns and everything gets enriched. When something doesn't, it now stands out like a sore thumb and goes, that is not the right way to do business. And then you look that and you say, translate that then into IT and then into data centers and there's these natural big red flags that says, that's an old way of doing things, that's the old economy, that doesn't enable me to go forward. I need to go forward, I need more agility. You know, I've got to get data freedom and then how do I solve that issue and then what companies are going to take me there because they're thinking the same ways as we are. This is why Tesla's screamingly successful. This is why something like these beds are there. This is why things like Phillips Hue systems are good and the list just goes on and on and on. We're naturally inclined to work with products that enable us to enrich our lives and actually give feedback and then benefit us over the air. We don't like things that are too static now. And actually there is this whole philosophy of cloud which I think from an economic standpoint is superb. Our product is tier one enterprise storage in an SDS fashion for public, private and hybrid clouds. But we're seeing a lot of people doing bring backs out of the cloud. There's a whole thread of it right now. But I would actually say maybe it's not because the cloud philosophy is right but it's the business model that the cloud guys have got because a lot of people have looked at cloud as a set and forget. Dump my stuff in the cloud, I get good economics. But what we're talking about now is data gets poked and prodded and moved and adjusted constantly. But the movement of the data is such that if you put it in the cloud it's going to impinge you based on the business model. So that whole thing is going to mature as well. Right. But you're in such a good position too because the growth of data is going bananas. We were just at RSA a couple of weeks ago and one of the conversations was about smart buildings. There's IP devices on shades that tie back to the HVAC and if anybody's in the room or not it should be open, should be closed. Where's the sun? But there was a really interesting comment about if you look at things from a software defined way you take what was an independent system that ran the elevator, an independent system that ran the HVAC, an independent system that ran the locks, one that ran the fire alarm. But guess what? If the fire alarm goes off maybe it would be convenient to unlock all the doors. Maybe it would convenient to automatically throw the elevator control system into fire mode which is don't move. Maybe, so in reconnecting these things in new and imaginative ways and then you tie it back to the IT side of the house it's getting a one plus one makes three effect with all these previously siloed systems that now can be connected. They can be software defined. You can kind of take the operation to a level. I would have never thought of that in a hundred years. I thought that was just again this fascinating twist of the lens and how to get more value out of the existing systems by adding some intelligence and adding this back and forth telemetry. Yeah, and again, part of me is being the CEO of Deityra I want to advocate it's the right platform for people to use but part of this is my visceral obsession of this market is moving through this software defined pattern. So it's going from being hardware resilient to software resilient to allow you to have flexibility across it but things have to kind of interconnect and work as you just described. An SDS software defined storage as an example comes in different forms. HCI is an example of it and clouds an example. I mean, everything is utterly software defined in Amazon. So it's the term gets misused. It could be software defined. You could say data centric, data defined or you could say software resilient but the whole point is what you've just described which is open it up, allow data freedom, allow access to it and then make sure that your business is agile and whatever you do can take the feedback in a continuous loop. And it allows you to move forward as opposed to I've just got this set and forget or lock mentality that allows me just to sort of look down the stack and say I've got the silo, I'm owning that customer I've owning the data and by the way that's the job it's going to do. So this is just, it's the whole concept of kind of people opening their eyes and my encouragement and I would encourage anybody whether customers or basically vendors is to look around your life and figure out what enriches it from a technology standpoint and odds on it will be something in the arena that we've just described. Do you think it's because I think software defined maybe in its early days was just kind of an alternative thought to somebody doing it, to flipping switches but as you said in the early example with the car propulsion it wasn't kind of a fundamentally different way to attack the problem it was just applying a different way to execute action. What we're talking about now is a totally higher order of magnitude because now you've got analytics you actually want to enable action based on the analytics, based on the data for your car to actually take action not just a guy maybe you should give an alert and notice it pops up on your phone so maybe we need something different because it's not just redoing what we did a different way it's actually elevating the whole interaction on a whole different kind of level. And this is kind of, and thank you for that it was the profound kind of ah ha I got wasn't joining Dateer and watching it I got a demo of the cloud UI the callback piece of what Dateer has and I was watching a dashboard of a live data stream of information that we were getting back from multiple customers and in each of the customers it would make recommendations of how many times it would hit cash and so it was actually coming back dynamically and recommending moving workloads across onto all flash systems you can do things where once you've got this freedom an application, a data set isn't unknown it's now basically in a template and you say this is what priority it has and so you say it's got high priority so whatever the best latency you can give me give me you drop it onto a disk and at the moment I've got hybrid that's all I've got but I decide to add all flash so I put some all flash into the system now it becomes part of this fabric and it spots it and goes well hold on a second that will service me better and then migrates the workload across onto it without you touching it so in other words complete lights out so that the whole thing of this is what Mark and the team have done is looked at and said the only way forward is running this massively agile data center based on a swarm of servers that will basically be plugged together into something that would look like a fabric array but you can't then you've got to assume that you can now handle application life cycles across onto it, it'll make recommendations like the bed thing, you know what I was saying I was lying there and what I liked about it is I set my thing to 59 and then it realizes I'm not sleeping very well it starts adjusting 60, 61, 60 sleeping well okay that's it and then that's good we'll do the same thing where an application will actually say here's my template this is what it looks like, it's top priority by the way I need the most expensive drives you've got drops it onto it and then it'll look at it and go actually we could do just as good a job if it was on hybrid and it'll migrate across and optimize the workloads and so it's not, again part of it is not day-tier is the best SDS and it is for tier one for enterprise storage it's the fact that the entire industry no matter where you look at it not just our industry but everybody who's providing tech is doing exactly the same thing which is and you kind of look at it and you go it's kind of edge asset micro telemetry and then that feedback loop and then continuous adjustment allows you to be successful that's what products are basically getting underpinned it's one of these trends I just I know we're almost out of time but I just can't help but say it because we used to make decisions based on samples of old data with samples and it was old and now because of where we are in the technology life cycle of drives and networks and CPUs and GPUs we can now make decisions based on all the data now and what a fundamentally different decision that's going to drive us to and then to your point it's like what are you optimizing for and you don't necessarily optimize for the same thing all the time that may be low priority workload optimized for cost and maybe a super high value workload optimized for speed and latency and that might change over time when a new workload comes in so it's such a different way to look at the world and it's temporal right I mean I know you're going to kick me off now but think about it right the old days of writing a car building a car is you thought well what's going to need to be in the car in three years time put it in now build manufacturer coming out and then with a Tesla I buy the car in December since December I've now got pin based authentication I've got Century mode I've got dash cam they've all gone all free I've got a pet mode into it now my car's got more range it's got higher performance it's got higher top speed and I haven't even taken the car into it it's all over the air and this is all about continuous optimization that they've done around the platform and you just go that's the way it is LinkedIn recently someone posted something and said you know QBRs are dead well the reason they're saying that isn't because there's a stupid thing to do QBRs it's because if you're not measuring your business and adjusting on a continuous basis you're going to be dead anyway so our whole economy is moving this way so you need an infrastructure architecture to support that but where everybody's the same we're all thinking the same and it doesn't matter what industry or you know proclivity you have this adjustment and this speed of adjustment is what you need and like I said that's why I wanted to get to Deityra that's why I'm excited about it and that is the aha that I had as I kind of looked at it I went oh my God I'm now working with cloud people who understand what they've walked in the shoes and I kind of got this weird sense of can you imagine what it would have been like if you were Elon the first time you saw 100,000 cars worth of live data spilling in of what power you have to adjust and to basically help your client base and you can't do that if you are in fixed things and so that's the world moving forward just in time for 2020 when we'll all have great insight in a few short months we'll all know everything well guy great to sit down love to keep tabs on you on Twitter and social and thanks for stopping by thanks Jeff appreciate it all right he's guy I'm Jeff you're watching theCUBE we're having a CUBE Conversation our Palo Alto Studios thanks for watching we'll see you next time