 Live from San Diego, California, it's theCUBE. Covering Cisco Live US 2019. Brought to you by Cisco and its ecosystem partners. Hey, welcome back to theCUBE's coverage. Day two of Cisco Live from San Diego. I'm Lisa Martin. Dave Vellante is my esteemed co-host and we're pleased to welcome one of Cisco and Cohesities customers from Quantium, Craig Taylor, executive manager of business technology and platforms, Craig, welcome to theCUBE. Thank you, it's great to be here. So, we love talking with customers, we love talking about data. Tell our audience a little bit about Quantium. I know you guys have expertise in two core domains, data science, AI, two really sexy topics that we talk about on theCUBE at every event. But give our audience a little bit of a flavor of who you guys are. Yeah, so Quantium's been around for 16 years, founded and headquartered in Sydney, Australia. And really there are, like you mentioned, the two main aspects of our business. So we think of data science more as human intelligence and then the AI side is how we can augment that with computers as much as possible. So on the human intelligence side, we're looking at things like data curation, how can we work with a company to understand their data, perhaps monetize their data. Then on the AI side, we're more looking at things like, how do we do predictive modeling or predictive analytics? And how can we get that in front of, maybe say a supply chain solution or working with grocery stores around actually predicting how much fresh food they need. So we think of these things like, wouldn't it be great if we had a better idea of how much we needed? Less waste, less cost, everything else. That's really how we kind of split the two sides of the company. And you guys provide this as a service, is that right? Yeah, that's correct. So with those two arms, we focus on whether it be a consulting engagement with a company, whether that's a one offer or an ongoing thing. And then we have a range of products that we sell as well with the idea that any of these companies, whether it be a bank or a retailer, can plug these tools into their existing solutions to give them some real data, but some real impact as opposed to the thoughts or the feels or the gut instincts that we've been working on for so long, right? So paint a picture of your environment. I mean, what does it look like? And cloud, not cloud, apps? It's only a variety. So if we think on-premise is really where we do a lot of our work. This is around a lot of companies still feel a little bit sensitive around where their data is going. They like that security of knowing physically where it's located. So on-premise stack, we have a bit over 300 servers running a Hadoop cluster. That's where we do the majority of our AI work. And then what we augment that with is, and what we use the cloud a lot for, as we're doing work globally, we're doing a lot of work in North America, it's not feasible to bring all that data back to Sydney, process it and send it all back. So then really what we use the cloud for is to take our technology, take our analytics through the data. So if we're working with a customer, West Coast, East Coast, and they're in Azure, we'll deploy in Azure. If they're in GCP, we can deploy in GCP. And that's really how we use cloud is to offer our service as much as we can around the world. So you said you got 300 servers that I hear you write in a Hadoop cluster, right? Yeah, correct. What's your distribution? So we use MapR at the moment. I know there's certainly been a bit of news about them. Well, all three of them. Yeah, so. Well, I guess Hortonworks now fall it in, but yeah, correct. It is, cloud has certainly shaken up that marketplace quite a bit. For sure, yeah. It's been something that we've been keeping a close eye on for quite a while. What's the future there? Is it another distribution? Will someone pick up MapR? Will they get through it? So it is interesting. It's certainly a challenge, but when you're playing in a more emerging space, these are some of the risks you take, but we've always felt that they're worth it. We've had many great years of that, and we don't really see any reason that we're not going to get more great years out of that Hadoop environment. Yeah, I mean, the IP is going to survive, and it sounds like you guys were early on into it. You got a lot of value out of it. If you had to do it again, you'd probably do the same thing. Yeah, I mean, that's certainly true. I think what we've built, you know, there are cloud options, you know, on the hyperscale providers that you can use, but look out of the box, they're not really capable of what we were trying to do. So if we had our time again, we probably would still build the same solution. We'd build it a little bit quicker, obviously, because it's a little bit more in the marketplace. It's not such an emerging technology, but I think we would do the same thing again. And MapR was always ahead of the game. Correct. We're there, we're approached. So obvious question is, how do you protect that data? You know, you're a Cohesity customer, but talk about sort of the data protection aspect of that. Yeah, and so this is, you know, where Cohesity really had a lot of synergies with us was centralizing a whole raft of data sets in the one location, and that's what we do with Hadoop. You know, we take a lot of different data sets, and you know, we put it all there, we aggregate it there. So on the secondary data side, we had the same problem, you know, siloed data sets all over the environment. You know, things like, you know, the protection aspect, the compliance aspect, it's not impossible, but it's very hard to manage. So what we really wanted to do was, what do we do with the data when we're not using it anymore, you know? So we might still want to use it in the future. We have to hold on to it, and we needed a better solution for how we managed that. So having Cohesity, which to us, you know, being a hyper-converged solution, it's very similar to how Hadoop works. You know, it's a lot of data, you know, a lot of compute, and that's how you deploy it. So we found that actually having all of that, the secondary kind of data that we still needed to keep combined into one location, for us, it matched on a technology level, and then being able to have all that data in one space, you can do some analytics on it, you know? How often are we using it? What is the data? How many copies of it do we have? So there are a lot of synergies from the data science aspect, and also the technology aspect that just worked really well for us. So, I mean, what was profound about Hadoop was the idea of bringing, you know, five megabytes of code to a petabyte of data, leaving the data where it is, highly distributed environment. Obviously, you know, challenge protecting that. Help us understand, you're saying that Cohesity architecture is well suited for that type of environment? Yeah, it certainly is. I mean, it augments it quite well, is sort of how I'd say. So at the moment, we keep the environments, you know, quite separate, but the way we manage them is very similar. So, you know, there's great, you know, audit logging, great security controls that you can place on both environments. So the way that we structure Hadoop with, you know, role-based access, who can perform what action, the same thing applies in Cohesity. So now we sort of see that the way that we manage primary is the same way that we can manage secondary. So it's easier for the staff, when we come to things like, you know, compliance or legislation or, you know, I mean, we value data, you know, it's our lifeblood, so we have to be very careful with it. So if we want to do any audit reports or anything like this, we can do them the same way. Who has access, what they've done. So Hadoop's been around a lot longer than Cohesity. So what were you doing before Cohesity and what were some of those challenges? Yeah, what we were doing was a lot. And that was really the only option we had. So we had four or five different solutions that had kind of organically grown over time, whether that was, you know, some secondary storage, multiple different backup products, throw a couple of nazis in there for, you know, just for good measure. Just in case. Yeah, just in case. And then really what we were doing and how we managed that is we had close to one FTE dedicated to that environment, you know, and it really, it's not great for that person. You know, it's not really the funnest of jobs. And then obviously the management of it could becomes quite difficult. And so that was the, that was how we did it. We got by, but it certainly could have been a lot better. So that was one FTE dedicated to the backup. Just dedicated to the backup. Data protection, okay. So you bring in Cohesity, you do the business case. Oh, wow. And part of that was we could free up this person to do other things, I presume, right? Yeah, definitely. That was actually certainly one of the key business cases. So, you know, IT is a cost center. We certainly, you know, we work for the business. We support the business. There's no doubt about that. But we are at the end of the day a cost center. So getting extra headcount or getting equipment, there has to be a really good business case behind this. And so we found that, so we freed up about 80% of time that we're spending on this. And so actually the two biggest things that we've seen as a benefit of that, staff engagement is actually a lot higher, right? Because we don't have someone just dedicated to, you know, turning the screws on this old solution all the time. So they get to spend more time on newer tech, which is great. And then obviously if their time's freed up, value-added activities, you know, what can they be focusing on? So how does it work? Is it sort of a self-service platform now or somebody, this individual, sets the overall policy and then people apply it as they see fit? We have a range, you know? So our infrastructure team holds the overall management of it. And we have that one person who kind of, you know, say rules it, so to speak. But the way we've done with this role-based access, you know, we can give the service desk permission to search backups. So if someone needs a, you know, needs a restore or, you know, maybe legal and the compliance team want to know, you know, who is accessing what, we can give a lot more self-service to these teams. So the service desk, if they're dealing with, you know, an end user that wants a restore, within 30 seconds we can tell them, okay, here is the backup we have, here are the dates that we have it, which one do you want? You know, previously that's a week and a half turnaround, you know, escalate the ticket, spend three days doing restores and searching through it. Working weekends. Right, working weekends. And if you even do have the data, typically what happens by the time you've restored it, the customer has said, look, well I don't need it anymore, it's too late. Yeah, right? So let's talk about some of the customer benefits. You've only deployed this about six months ago. You talked about a number of the benefits from a time perspective, allowing valuable FTEs to not only be reallocated for other projects, but also from a job satisfaction perspective, which was all the way up to the top line of the business. But in terms of helping customers extract more value from their data, monetizing their data, that example that you just gave, we're too, too long to recover data before and the customer, the time has passed. What are some of the impacts that your customers are achieving so far? Yeah, so I think one of the biggest area of this that I think we actually look at the most is that, you know, like I mentioned earlier, we will do say a piece of work with a customer and then we'll keep that data. We might need it in the future, but you know, there's not an ongoing engagement. You know, what are we going to do with that? And so we tend to sort of put it aside. If a customer wants any further work done or perhaps they want to come back with clarification or anything like this, it then takes us quite a bit of time to find that data, get it back into production, you know, get it back to the state that we, you know, we were previously using it in. So one of the biggest things that we've seen is actually now having all of that data always available on Cohesity and being a hyperconverged platform, it has a lot of compute on it as well. So we can actually run some simple analytics on that data. So if a customer comes back and wants to query just a couple of small items or perhaps we want to recheck a couple of things, super easy now for us to do that. And so we talk about, you know, time to market. You know, anything like this is really big for us and customer responsiveness. So if a customer is asking us a question and the answer is a five minute answer, they don't want it in four days. So if we can turn that answer around a lot quicker, then obviously everyone's happier. And you've already been able to sort of achieving that. Yeah, we have been able to start achieving that already, whether that be from a customer perspective or certainly from a compliance perspective. You know, if we have a customer that actually wants to know where is our data, who has access to everything else, we can turn that around straight away. So obviously when we talk about customer satisfaction or that relationship, they feel a lot more comfortable that we're doing the right thing with their data. And that is obviously hugely invaluable as for us as a business. And just another infrastructure question. These 300 servers, it's mostly UCS, is that right? Or a lot of UCS? Yeah, so we use Cisco for pretty much everything. We certainly are heavy users of UCS. And so when we're looking at, I mean implementing anything to the environment, you don't want it to be a lengthy process because your return on investment is going to be hit. If you're spending three months installing something, you've already paid, you're getting no benefit out of it. You know, it's now three months old before it's even implemented. So having this kit on Cisco UCS has been great for us. And we were having issues with our previous backup solution, and we actually managed to implement the Cohesity solution on UCS and start using it before repairing our existing solution. So it's phenomenal how quickly through UCS we're able to bring it in. What kind of issues were you having? Just integration issues? Yeah, so with our previous backup solution being a fragmented solution that we had stitched together, we had something as simple as a rate control of failure caused a whole bunch of data corruption across multiple areas. And so how the NAS saw the data corruption was different to how the sand saw it and trying to re-index everything. We're struggling to understand what was going on. And whilst we're working through that, we actually had some other members of the team implement Cohesity and get it into the environment quicker than we could repair our existing solution. And that's the power of Cisco UCS, really. Looking at this massive transformation that Cisco has been undergoing for a while from a traditional network appliance vendor to hardware, software, what are your thoughts on how that transformation, which is in part you could say accelerated by DevNet, how is it going to enable businesses like yours to be able to start getting value even faster from the technology? Yeah, that's a very good question. And that's something I think a few of us in the industry, if we go back sort of two, three, four, five years, was Cisco going to reinvent itself? What was their place with hyperscale cloud, these kind of things? I think quite a few people had some questions around what was going to happen in that space. You know, they weren't always the quickest to market. They had great products, but there was a bit of speed issues there. And what we've seen as they reinvented themselves is, you know, Cisco has this great name for really, you know, being ahead of the curve or leading industry. And this is, I think, what they were built on, really. And so it's been great from our perspective to see them say almost getting back to their roots a little bit in this regard. And so for us, you know, we are a technology business. We are fast moving. Our customers want things to be fast moving. And so being able to rely on a technology partner like Cisco and knowing that they're looking for the latest and greatest even quicker than ourselves, I think that's probably where we start to see the biggest impact. You know, in the past, we might have a challenge that we need to solve. You talk to some vendors and you might hear something like, oh, you know, we're working on that. Maybe in 12 to 18 months, we'll have it in the marketplace, but we need it now. You know, we don't need it in 18 months. It's a today problem. And that's not what we're seeing anymore with Cisco. Typically, any conversation we have with our count reps around, here are some of the challenges here, what our customers want to do. More frequently than not, you know, our Cisco count reps will say, I think we have a solution for that. And that really, I mean, being able to partner with players like that in the industry, that makes some of the biggest differences for us as a company because we need to partner with all these people to do what we do. Exactly. So with all the momentum that you guys have achieved in just six short months, what's next? Yeah, and this is, Quantum is certainly a fast moving company, like I mentioned. And, you know, what we wanted, we always like to run close to the leading edge. We're similar with the Duke. We like to be early adopters. We like technology to grow with us, right? And this is what we saw in Cohesity. So they haven't been around for long and they're already doing everything we need. So we think, well, this is a great mix. You know, if we've got someone who's already solving everything that we need, this question of what next is great. And so as we move more towards, you know, your hyperscale cloud, being able to run Cohesity across all those environments to manage all of that data across all of it, that's certainly a big one that we're investigating. Like I mentioned, we keep pretty much all of our data, you know, and so actually being able to use cloud as an archive solution, it sounds great, but then it's another silo to manage. It's, you know, it's another solution that you need to implement, but Cohesity will manage all that for us. So the what next I think is we'll see the, we'll see the scale out of the solution as our data, you know, requirement grows. We will see it expand into the cloud environments that we're going to start building. So we really see it growing with us from that aspect. And then we see a great idea of being able to repurpose a lot of our on-premise hardware by archiving out to the cloud as well. What about SaaS? You see a need to use a Cohesity to protect your SaaS data or kind of not there yet? Yeah, I think it certainly has a play there. It's still something that I think we're exploring a little bit more to make sure that it's the right fit, but certainly there is an opportunity there to be explored. Always opportunities. Well Craig, we appreciate you stopping by the Cube and sharing how Quantum is leveraging your partnerships with Cisco, with Cohesity to drive those core business drivers of data science and AI. Thank you. Our pleasure. For Dave Vellante, I'm Lisa Martin. You're watching the Cube Live from Cisco Live, San Diego.