>> Live from Las Vegas. It's theCUBE covering AWSre:Invent 2018. Brought to you by, the Amazon Web Services, Intel and their ecosystem partners. >> Well, welcome back here to the stands. As we continue our coverage here on theCUBE of AWSre:Invent as the day starts to wind down here. Still a lot of energy out there on that show floor. As we are packed with all kinds of great exhibits here. A lot of interesting folks here still making themselves at home. Tom Murphy's with us now, along with Justin Moore and John Walls. He's the Chief Marketing Officer, Turbonomic and Tom glad to have you here on theCUBE. Thank you for joining us. >> Yeah, great to be here. >> So just tell us a little bit about Turbonomic, first off. And then we'll drill down a little bit there. So why you're here at AWS but what do you do for the folks at home? >> Yeah, ultimately what we're doing is we have workload automation for hybrid cloud. And workload automation for us is ultimately where we go out, we discover the workloads, we optimize performance, cost and compliance simultaneously in real time. So what that gives for the customer is we call them smart workloads. Self managing anywhere in real time. The outcome of that for the customers is, they get guaranteed performance, a short performance, a short compliance. And they eliminate a lot of the complexity that they're experiencing today. >> So you're trying to grease the skids, more or less, right? >> Grease the skids, make sure that they're life's easier and they can actually accomplish the outcomes they want. >> Complexity's been a theme that we've been covering in the last couple of days. It's come up quite a bit. The customers are struggling with the amount of choices. We had Andy Jassy on stage today, again, announcing another zillion products today that the US has created. And that gives you a lot of flexibility. It means that you can optimize for particular choices that suit you very, very well. But being able to choose between them can be a pretty daunting task. Now how does Turbonomic help customer decide which of these choices is right for them? >> What we see from our customers is that they're actually looking at typically three platforms. They're still running on prem with VMware. They're looking at AWS, of course. That's why we're out here. And they're looking at Azure as well. So when it comes to choices, they want that flexibility to decide where the workloads can run. By looking at the workloads versus the infrastructure, we abstract the work that's running. And we can model for them, for example, a VM workload that's running on pram, we can pick it up and say what would that look like if it runs on AWS? What will that look like when it runs on Azure? So for us, by abstracting the work from the underlying infrastructure, that gives the customers the flexibility, the simplicity to understand and de risk any migration projects that they have. >> When you identify the way something could go based on what the workload is, do you just tell customers what that is or are you able to automate that decision making process for them? >> When a customer decides to actually deploy workloads they don't know necessarily how big the resources should be. They don't know where should they be placed. But our analytics engine, what it does is it goes out and says we understand the complete infrastructure, we're agent less, we plug into vCenter, we plug into Azure, we plug into AWS, we pull the configurations back and when we decide where to place it, we know best where to place it. We decide how big it is, we know best how big it should be at that time. Should the demand change, what we can do is actually dynamically adjust, in real time, automatically, the size of the workload, where it lives. We can scale out on prem, we can multiply the number of instances or we can actually scale up which means add more resources to the actual workloads. >> And what about that decision then because you talk about on prem and a lot of people have heard a lot. Making the move over, going public. There's a little bit of pain there for some people, right? >> For sure. >> There's some barriers there, what are you saying and what are you trying I guess to lead people through that consideration so they get it. Alright, so it's going to hurt a little bit, maybe. >> Yeah, for sure. >> But, ya know, we're going to be a lot better off by the end of the day. >> Ultimately what we see is, especially on a lift and shift where customers take existing workloads and they move them to the Cloud. A lot of times when you think about how utilized they are on prem, they tend around 50% utilization. So if you actually take that box of, lets say the resources that you've defined on prem and just pick it up and drop it into the Cloud, you're 50% over prevision. Part of the pain point is knowing exactly the resources they need, but understanding not just what you allocated but what you consume, which is a smaller view. Looking at the consumption and using consumption in the Cloud versus the allocation. That's a quick, efficient use case for making sure that they use exactly what they need when they get there. But then not compromising performance when they get there. They have the right performance at the right cost. >> This isn't a static decision either because as we've seen, there's new announcements everyday so we get new instant types that have been announced at the show, but also workloads and the demands for what customers need to do with those workloads. It's constantly changing, so you need to be able to react to that and to change what the right option is from moment to moment. >> And then when you add on top of that, like reserved instances, right. So the complexity of what instance type to use, let's say there's millions of choices when you look at the combinations. Ultimately, there's new one's introduced regularly, so how do I take advantage of that? There's discounting that's applied, there's discounts that come out, bundles that come out. And also the RI's. So RI's, thing of the two metrics for RI's, one would be utilization of the RI's, so I want to make sure if I actually buy RI's and invest in RI's, I want to make sure I use them. And the second is, out of all the instances that I have, what percentage of my environment actually is RI's? So think of that as coverage. So the two metrics that we look at closely with RI's as an example is coverage, which means I'm taking advantage of RI's, I'm not just doing one percent, I'm doing more than that. And the second is utilization which means of the one's I've purchased, I'm making sure that I'm using those, they're not going to waste. >> So customers who are doing this well, like clearly you've got plenty of customers who've done this successfully. So what is doing this well look like? >> Sure, well they start with an assessment, which they look at their on prem in many cases. They right size that, they run models, they run plans as to when they're looking at workloads that are picking up and moving to the Cloud, what do I need when I get there? Once they're in the Cloud, that's just the beginning of the journey, really. And what it is, is they continually optimize. And the continual optimization means that we're constantly meeting, I started talking about supply and demand earlier. But constantly making sure that the demand of the workloads is matched with the underlying supply at all times for the benefit of performance, for cost and also making sure we're compliant with business policies at all times as well. >> So if you have a customer that maybe comes to the show, and the catch the bug, right, they got the fever. We're going to go back, hey Tom. They're on the phone tomorrow, we got to go now. Do you ever have to tell people, just slow your roll, we're going to do this in a methodical way, we're going to do this in a responsible way. We're not going to go nuts. I know you want to go -- >> Go now! >> But people get excited, right? >> For sure. >> So, I mean, how do you handle that? >> Well, I think what I hear from customers today is you guys talked about complexity, right? So, if you think about complexity, and then on top of that you think about a little bit of a skills gap that's happening because there's really not the maturity of the expertise to manage a lot of the transitions that are taking place. And then lastly, once they're in the Cloud, again, I said it's not done, right? So to address really that how do they get through that, what do they use? We don't necessarily say slow down because we can actually get people to the Cloud very very quickly, in a responsible way. The thing that we like to say is we take the guess work out so what we're doing is we're taking the analytics, we're giving them intelligence that they can actually make very rapid decisions. Our solution can probably make decisions about what to do faster that people already to make the progress so, ultimately, we want to go at the pace they want to go. We've had customers call us on a weekend to ploy the software, actually go live like in a couple days. So, it's up to the customer. We feel confident in our decision making. You can't automate decisions and actions if you don't feel confident about it and we've got the customer points that give us to prove. >> So based on what you've seen so far at the show and your experience with customer who have been moving to Cloud, figuring out where they're going to put these workloads, what's next? What do you think people are going to be doing next? >> It's a great question, 'cause as a company, I'm really proud that we started as VMturbo many people still know us as Vmturbo. And what happened was Virtual Machines was were we started. We plugged into vCenter, we pulled information back and all of the sudden we're making decisions and actions about what to do on a virtualized environment on prem. Well, we started doing the Cloud, all of the sudden its like well it's more than just VM, it's Cloud too. We literally had to change the name of the company to accommodate the capabilities. So by having this economic supply and demand model, it'll alow us to really just apply it to not just VM on prem, not just Cloud. So to answer your question, containers or micro services is steam rolling into, we hear that in all of our customers now. When I came here three years ago it was worth thinking about Cloud. Last year, actually, we're testing, right? This year it's we're live. Next year it's going to be containers, containers, that's what I think cause containers are just going to be just coming in, Cloud Native applications, we're past a litte bit of lift and shift. We're moving into Cloud Native so that's what I think is going to happen next year. >> I think you can stick with Turbonomic. I think you're okay for a while. >> (laughing) Sure. >> Hey thanks for being with us Tom. >> Absolutely guys, I appreciate the opportunity, thank you. >> You bet, have a great show the rest of the time here in Las Vegas. >> Thanks very much. >> Excellent, Tom Murphy joining us from Turbonomic and that will be it for this day here on theCUBE AWS:reInvent. Back with you tomorrow Thursday for Justin Moore and I'm John Walls, thank you for watching. (electric sounding music)