TEC37 E04: AIOps and the Current State of Today's Technology

Please view transcript below:

Robb Boyd:         AIOps is the application of artificial intelligence for IT operations. It's the future of IT operations combining algorithmic and human intelligence to provide full visibility into the state and performance of the IT systems that our businesses rely on. This is where things are going, and it is critical that this type of thing is part of your future when it comes to digital transformation. It can seem like a big intractable problem, but we've got a lot of experts here to help us. Welcome to TEC37, the technology education collaboration podcast. It's part two of our three part series on AIOps and APM.

                        Well, gentlemen, welcome. Here we go. Get started on part two of three AIOps and APM, and go a little bit deeper down the rabbit hole, and we've got a couple of new faces joining us today. Arsalan from World Wide Technology, appreciate you joining us today.

Arsalan:            Thank you so much. It's a pleasure.

Robb Boyd:         What is your role with, with World Wide?

Arsalan:            So I am the technical architect down at the APM and AIOps practice, working with Tanner.

Robb Boyd:         Okay. So you get into some details. That's going to be important here as well. Now Ben, I want to welcome you as well. You are with AppDynamics, is that correct?

Ben:                 That's correct. That's correct. I help support WWT in their APM and their AIOps practice. I'm just helping build strategy and looking at the market and figuring out how we can bring a solution to the clients.

Robb Boyd:         Yeah. And what you guys have done together is something I'm looking forward to getting into further today. And Tanner, for anyone that watched part one, you would hopefully recognize Tanner from part one. So Tanner, I think you're actually perfect. You're with World Wide Technology. I'll have you please tell us what you do for World Wide and then maybe bridge us, if you will, from what we covered in part one and how that's setting us up for what we're covering here today with part two.

Tanner:            Sure. I'm the director of application performance monitoring in AIOps for World Wide Technology. So, it is my job really to bring to bear where we stand and what we can offer and deliver inside of both of those spaces. And I have the distinct pleasure of working here with Arsalan and Ben to make that kind of vision a reality. So, to bridge us from where we were to where we are, our initial intention with this series was to really take and distill what we've learned for our customer base. For our partners and our customers, and take the concept of AIOps, which is both complex and quite a complicated. We've got quite a few different components and moving parts.

                        And so the series itself, the first one was really designed to create or define the executive messaging in that space. Help us understand from the top down. And then the second part of this series, really to start talking more about the technology. What comprises it, what components, what architecture, and really the focus of these two gentlemen here as to where they've been working and what they've been building out.

Robb Boyd:         Yeah. One thing I think that's really interesting is, as I understand it, and you tell me who's best to maybe speak to this at first, but as we put a little bit more meat on the bone, you guys have personally, really taken it on from World Wide Technology, building out and defining AIOps and saying, we need to actually use the tools that supposedly will answer certain questions that help fill this out because it's moving from concept to reality. I don't know, who's best to speak to what you guys have been building towards and how that fits into how we're going to learn from you in that process today?

Tanner:            I can start.

Robb Boyd:         Ah. Thank you.

Tanner:            I can start at least with the perspective and Ben and Arsalan could talk about what we've actually created. Our intention was to take and really create a comprehensive ecosystem around artificial-intelligence-driven IT operations. Really what that means when, about a year ago, or a little more, I guess. Actually about precisely year ago, we were walking around Cisco line and a lot of folks had AIOps in their messaging. A lot of folks were talking about what AI based it operations look like and what it was comprised of. And I had a hard time finding a real functional architecture for that. And so I came back home and just set out with our team and our partners here at AppD, to figure out what that looked like. And I'll turn that to Ben or Arsalan, you guys can jump in there as to really where we went from there.

Robb Boyd:         Well, you mentioned AppD so Ben, why don't you jump in and tell us where AppD fits into... Well, let's start it with this point, because one of the points that was made in part one was the fact that AIOps should not be thought of as, "Hey, go buy this tool." And it's AIOps in a box or even application performance management. But I know AppD does a very good job, probably filling in a lot of blanks that are important in this situation. But I wonder if you could give us some more intelligent definition of what AppD's responsible for and where that started to map into what you guys have been working on together.

Ben:                 Yeah. So as Tanner had pointed out, there were a lot of vendors out there, a lot of products out there that have AIOps, but when it comes to an actual architecture that covers IT operations, you'll find a lot of those AIOps vendors or tools or tool sets out there cover certain personas, but it doesn't cover the entire IT operation for especially enterprise clients. So when we first started looking at this, we said, yeah, we can come at this from the APM space. And we can look at a bunch of different things from the application, but we're going to have to work with other tool sets out there in order to cover the entire personas in the IT operations.

                        And so we decided to start with an application view of everything. And so that's where I AppD comes in. We put that application view across the enterprise, but it takes... It's really one of those, it takes a village to look at everything and hit all those personas and keep them on the same track. So that was the vision. Tanner was really instrumental in coming and saying, "Hey, it's great. You can do it in this space." But he wanted the whole pie. He went to cover the whole enterprise. And I think that's what we're doing.

Robb Boyd:         Well, Tanner has described himself to me, and probably doesn't want me to say this out loud, but Arsalan, he tells me that you do most of the work in terms of the technical reality and building stuff out. Taking stuff from theory to practicality. What have you been involved with and what has it taught us that's important for as we put more meat on this bone around this?

Arsalan:            So it's kind of important. For me it was important for me to figure out why we're doing this, right? So I come from a developer background, I've worked with fortune 50 companies in the past as a developer, as a dev ops engineer. So I've progressed throughout these companies. And as my role, I kind of understood, okay, what was the problem then? And how would it make my past careers, how would it make those positions easier now? So I took that approach. Okay, APM is helpful, but so it's networking. So is databases. Everything doesn't tie in together. I've had those issues where we sat into war rooms, where we sat in for pretty much 24 hours. And there was no resolution in the end. There was 50 people on the call. We would have this constantly. So I took that approach. I was like, "Okay, what would make my life better?" And that's essentially the approach I took. So we look at a specific product, how that's going to help for a specific aspect of a dev ops engineer or someone who's in help desk, sitting in a call center.

Robb Boyd:         Yeah. Because when we peel that back, we're really looking at things. And it seems, it seems funny that we're having to work our way backwards into this, because from an industry perspective, there's so many elements, as you mentioned there, that come to play, but ultimately what it all boils down to, is what is the experience of the end user that we need to be as productive. And that means that applications have to be available. And the orchestra that happens behind the scenes to make all that happen is part of what you guys are saying.

                        We need better visibility too, I guess? And knobs to then be able to adjust to that. And of course, using artificial intelligence, I guess you're saying also how can we let the machines do more of the workforce so that we could possibly be focused on higher order bits, so to speak. So Tanner, what would you state is the state of AIOps today from a technical perspective? Is this a problem? Well, are we talking about stuff that can be done today versus things that could be done? And how do we start going deeper on that?

Tanner:            That's a good question. So a lot of this, and we're guilty of it too, because it's such a unique concept, we tend to float out into ether a little bit and talk about where it's going and what we can do with it, and the advancements that we'll make in the IT industry. But the reality is that you can do it now. So if you think about what AIOps is, it's isolating high value decision-making, right? Because as we've grown technology, what we've done is we just added more layers. We split up where processing and where data lives. And we split up where you may have authentication, you may have services that run in an application that live in completely different physical ecosystems that are no longer connected through the core elements. They're serverless or they're out in the cloud.

                        And so all of these things have made it very difficult to have a real high value relationship with your application. And when I say that, I mean engineers and technical architects and network engineers, and getting them to really be able to make smart human decisions. Arsalan just highlighted a point. You get stuck in a situation where, when there is failure, you don't know where it came from. And so APM has approached that broadly from the application centric side of the house. What AIOps does is it actually opens up all of the elements. So if you think the whole data center, everything we invest in, drives the performance of an application.

                        Network, data, database structure, data infrastructure, the cloud. Every component of what we build, AIOps can be built today. And all AIOps really is. And that's our approach to it. We kind of went a simpler route. Not a simpler route, a more, done is better than perfect, approach, in my opinion. Because one, I don't know... I'm a software engineer from a long time ago. I've not done it for a long time, but the concept still remains, what are we trying to deliver, and for what reason? Technology is the answer to a challenge, not there for its own purpose and own wellbeing.

                        So what we've done is we just started to take elements that were there. APM. So what else can we see? If we added other data to what we know about what AppDynamics is providing us, can we make more intelligent decisions? Can we make more automated decisions, if we can add more data to it? And then starting to look at that from the vantage point of all of the different professionals within the application infrastructure. So, it has been an exercise for us to say, this is the big envelope that we're aiming for, but starting today, instead of just talking all the time, which is what we tend to do, especially on these, I tend to talk all the time.

Robb Boyd:         Hey. Yeah, some of us like to keep talking.

Tanner:            yeah, right? Is really to deliberate. Say, let's take AppD, Let's tie it into Moogsoft. Let's take AppD, let's tie it into Turbonomic. Let's create individual patterns and connections that begin to give us the ability to create an AIOps infrastructure, and then build upon that.

Robb Boyd:         Well beyond the fact that I think everybody should be calling World Wide Technology for assistance with these things, because this is... You guys tend to do very well with the multi-vendor, seemingly intractable type of problems that that can kind of befuddle customers, because it's difficult to know where to take that first step. But if we take aside the fact that that to me, is the logical first step, Ben, where do you see most people getting with something like this? If they acknowledge that this is something big, but you got to start somewhere, what do you generally recommend? What, do they look around at their infrastructure? What kind of stuff should they be measuring right off the bat? What kind of advice would you provide there?

Ben:                 So, for fear of being narcissistic going to this, I would say, start with APM. And the reason I say-

Tanner:            I'll back you up on that. Yeah, that's fair.

Ben:                 So I've lived in the APM space through different companies. The reason I say that is that's how everybody, especially when you see some of the things that have gone on recently with COVID-19 and lockdowns, where businesses really do business through their application right now. You're seeing a lot of your traditional businesses that are really struggling going out of business. And then you see other businesses. If you look at Walmart, that was a traditional brick and mortar store, they've done a great job of transitioning into the digital space and being able to offer, not only like pick up at store, but home deliveries. And you're seeing a lot of those companies start to do that.

                        So the first thing is to get that application in, because that's the lens that you're going to look at everything else through. And then from there, it's just taking the pieces you already have in place and tying those in and putting that application lens over top of it. We're not doing rocket science or anything. We're literally taking every bit of metric we're getting and saying, "Let's look at that through an application and end user lens, and then make a decision based on that."

Robb Boyd:         Well, I want you guys to continue to jump in on this, no matter who I ask the next question of, but I'm going to go to Arsalan, because I'm just working my way around on this a little bit, because you guys are smart in a lot of these areas. And so sometimes it's difficult to know exactly what to draw out of you. But Arsalan, as you mentioned, you've got experience from the customer side. And I didn't know... Does that entail advice on saying you need to look at maybe your most important app? Is it as logical as look at your most important applications first and start looking at what data is available from those types of things or what's your opinion on where people would get started here?

Arsalan:            So, yeah. And that's basically when I first worked for one of the Fortune 500 companies or Fortune 50 companies, APM was the first craft we actually took was with AppDynamics actually. We essentially thought about taking an approach on picking the smaller applications, but there was no value out of it, right? So we picked up the biggest of the baddest app that we had available. That if it did go down, there will be billions of dollars of revenue loss. So, that's the first approach we took and it worked out really well. From that point, we the scaled to smaller apps and not as critical. And then we moved on to a different organizations. So I think that's a really good pick. It works well. And so far there's has been true value behind it.

Robb Boyd:         Well, what type of information do you first start getting back where you go, that's valuable and that's proving the time that was spent this. And this, not necessarily even from your own experience, but I know you all work with a lot of different customers. Are there some typical use cases where people are going, "Wow, I didn't know that I could maybe get a handle on that, because also what you're talking about is even though we may start with specific applications, I assume we're talking about information that's coming from multiple places that's being brought together for the first time potentially. Is there a common reaction or a level of surprise from anyone in terms of what they've started to see and react to?

Arsalan:            Great. I think from what I've seen so far has been problems. Like applications that they already know they have problems with and they have no idea on how to fix it. [inaudible] I think that's been one of the biggest ones. Right. Exactly. It's a Pandora's box, right? Once we install AppDynamics on a third party solution, that's that's not in-house, then you have code visibility that you'd never had before.

Robb Boyd:         Yeah Tanner, you're nodding your head. What are your thoughts?

                        Tanner, can you hear me? I'm just curious your thoughts. You were nodding your head there, I thought?

Ben:                 Tanner, you're on mute.

Robb Boyd:         Ah. [crosstalk] Welcome back, Tanner.

Tanner:            Yeah. I was going to say, I have to back up then on this. APM is key. You have to understand, it's as much cultural as it is technical in this first stages and first steps, because figuring out how the customer experiences is all that technology is pretty critical. And then being able to walk down that path with them requires, sometimes the willingness to realize that some applications might be broken in ways that have been chronic and need to be fixed. And sometimes you got... That's why when we talked, I think we mentioned this on the last podcast, is when we talk to the executive leadership inside of companies, one of my first questions is, do you have the courage to do this? Because this may mean calling out, shining some light in some places that people really don't want to chime into. To say, we have problems here, but this money is now connected to this technology. We can see by tuning both of those, what happens to each?

Robb Boyd:         Yeah. I think what you also said is you can't improve anything that you can't measure. You've got to bring it Visibility to things.

Arsalan:            You can't that you can't change what you can't measure. Yeah, absolutely.

Robb Boyd:         I always mess that up.

Ben:                 When I talk to clients, I like to talk about and compare it to cooking. So, I'm a food guy. I love food. And yeah, you go back to older applications, when we're first getting applications, you can have one guy that would come in and make an application. He'd do a front end, he'd do the application, he'd do a database in the back. Pretty simple. It was like mom's cooking. You could go in. Mom can tell you everything that was in your meal. But with the advent of agile, with so many different technologies, development has moved into almost like a fast food mentality. You have one guy doing fries and one guy doing the hamburger. One guy putting ketchup and mustard. One guy, pickles.

                        Problem is that everybody's using their own tool sets. They're looking at everything their own way. So you may have a client who says, I just want a cheeseburger, but you've got a guy who's throwing jalapeno cheese on there. You've got another guy who's got his sweet pickles. And someone's like, "Oh, Sriracha is the flavor of the month." So I'm using and catch up. And by the time it gets out to the client, he's like, "This is nothing what I wanted, this is a mess." So being able to coordinate all of that together is key.

Robb Boyd:         Well, an I... Go ahead, Tanner.

Tanner:            I was just going to add that's really where... Not to play a narcissist, but I'm totally going to do it. That is where the ATC is the biggest step for us.

Robb Boyd:         This is the Advanced Technology Center?

Tanner:            We're able to do that. [crosstalk] Absolutely. World Wide Technologies.

Robb Boyd:         Okay.

Tanner:            Yep. Advances Technology Center, which is kind of unique to us because it is all R and D. So we're able to take and pull in all of those different flavors. So if you think about AIOps consists of multiple points of architecture. You're doing network performance monitoring, you're doing resource management, you're doing application performance monitoring and management. All of those things have to live independently and to Ben's metaphor, as we're cooking each of those, the cool part about the ATC is we have test kitchens for each, right? So we're working on cheese on one place, we're working on the hamburger in another, and we narrow it down to the most effective way to create each of those elements. And the ATC blended, with their expertise allows us to say, instead of taking a guess at all of these things... So people come to us, they're going to watch this and they're going to say, "Well, why you? What are you creating that's unique in this space?"

                        It's that we have refined all of these key technical elements. Performance monitoring for the application. We've narrowed down and worked with that intensely at an executive and enterprise level. We've done it with network performance monitoring. We've done it with workload optimization. We've done it with cloud management, public and private. And so for us, what we're doing is we're turning that codecs to lock up the right pieces into what we see as not just our AIOps architecture, but our best-practice, best-of-breed, AIOps architecture, based on a ton of research and development inside the ATC.

Robb Boyd:         Well, and also to be fair, and I'll ask you this question, so I wouldn't ask Ben, but Arsalan if I weigh in on this. You mentioned a couple of other tools that weren't AppD. Moogsoft and something else. I'm just curious what led you to what tools that you guys have been building out in the ATC. What are these tools given you that you couldn't get elsewhere and how do how are they complimentary and not competitive? Because I feel like I'm always a little bit edgy about us just throwing more tools at things and always saying, "Oh, this is what you need to buy now." I know that's not the truth, but that's also, that's the easiest thing to justify sometimes when we're trying to get change going.

Tanner:            I think the tool question is a question that everybody should answer to some degree, because I think what it is, it's compositive of a lot of different things. I mentioned two players, but that's not, by any means limited. We have a very broad partnership organization that allows us to be the best-of-breed. To take the ones that our customers are bringing to the table and because AIOps is different than a pure play situation like APM. You can pick a vendor, you can pick a partner. When you're talking about AIOps, you're looking at the whole ecosystem. So just because I may have Turbonomic and AppD, I may want to use BigPanda. We may say, "Well, we want to partner with Moogsoft in this space." Or they may say, "We want to use this particular partner."

                        The difficulty in there is that you don't want to dictate the structure. I think I said this last time too, but people say, "We want you to be agnostic." And I always tell them, "No, you want us to be polytheistic." Right? You want us to have multiple expert tool sets. You don't want us just to go willy-nilly and pick anybody, right? So we narrow it down to key players. We partner with strategic partners. We work across the board. AIOps is a little broader in our partnership management structure, but all of that has come from customer request individually, best case scenarios, best integration. Sometimes it's what they can do that leads us to select them. And Arsalan, you jump in, man. You got the depth on this.

Arsalan:            Yeah. And then the one big thing I was going to talk about here is, we've seen integrations from all kinds of partners that like, "Hey, we are beginning to integrate with AppD." "Okay, show us what you got, right?" So at that point we're seeing, okay, they do have integrations, but there's no really true value that we're seeing. I want a specific thing out of you guys. If you can't provide that, then it's useless. If we were focusing on application centric, right. In that case, we're talking about AppDynamics. And if they don't have integrations with them, then we haven't looked at them again until they provide us something useful.

Robb Boyd:         What is an example. [crosstalk] Arsalan, I don't know if it makes sense, and I don't mean to force you into something, because I know there's always going to be overlap, but some of these other tools that you're using, what do you primarily think of when... BigPanda, Moogsoft? What is it, and it's not to say that AppD doesn't do these kinds of things or anything, but each one gives you certain answers you're looking for. And there's probably many others that you consider and maybe use for certain things. But can you give a little snapshot of what kind of stuff you're getting out of these tools?

Arsalan:            Sure. It's mainly how robust our API is. If we can get metrics out of their product, and it's more robust than what they're exposing, then I can take that data, combine it with AppDynamics or combine it with Turbo, and then we will get some really good value out of it. It's mainly how exposed our API is. Some of them have natural robust, but some of them are full blown, everything's exposed to the public.

Robb Boyd:         Well, so where in this scenario, if you've got multiple things that are all coming together, what is a customer in an ideal environment... Let's say they've been through some of or all of this journey. These things never end actually, but let's they're advancing on this they're reaching that Nirvana that we're promising here, where you've got artificial intelligence weighing in, handling some mundane stuff in there, but they've got visibility. They know why. They maybe even getting proactive on how things are responding or would respond in certain situations.

                        What are they looking at? Is there a central... Because it sounds, whenever I hear multiple tools, I always have a little bit of a fear of, "Oh, it's just more to potentially try and keep up with." And everybody's already going to be constrained a bit. I'm just kind of curious, how does it begin to all come together and where does that fit?

Arsalan:            Well, so we're talking about a centralized dashboard, right? Instead of you looking at 10 different tools that you might have in your portfolio. You're looking at one single pane of glass. I can choose single pane of glass that we have the entry points from the different products we just talked about, and then some, if you have existing, additional ones, all pointing to one place. That way you have that single pane. And then if you do want to bounce out to the specific product for that specific troubleshooting that we're looking at, you're more than welcome to, but everything's supposed to come together all in one place. That's your Nirvana, I think.

Robb Boyd:         And that's part of what it sounds like you guys have ruled out, is an ATC as you're working through this is you're going through the trial and error process. The customers probably wouldn't even have time to take this far in saying, is this exposing enough information to work together so that we can do the central dashboard? And even though under the covers, it may be a bit more complex in the multiple tools, but you guys have already been testing that out and throwing it up against a wall. Yeah.

Arsalan:            And coming from experience from working for top companies, I can definitely say that this is something I've definitely talked to with several engineers. These companies do already. This is something we wanted, but nobody want to take the initiative internally either.

Tanner:            So this is one of the things that we're doing here and to go back to a question you just asked about what data are we getting. The big differences that we're looking at this in the way that people use the data. So imagine you're driving a car, you get information on the tire pressure. That's a monitoring tool, right? That's a specific aspect of monitoring, but in all reality, do I care about the actual tire pressure or do I care about my safety or my comfort? And those are the two things that I care most about. So what if I could combine other data? Air conditioning health, tire health, seatbelt... Whatever it is. All of the elements that provided comfort, that's my dashboard. And I want to know how those correlate and how that is measured. It's almost like we're looking at AIOps on the other side of the fence. IT has come up with tools to monitor IT.

                        We monitor the network, we monitor the app, we monitor their data in our cloud, but what are we really trying to do? We're trying to make sure that app is up and running or that app is making money most effectively or that our key customers are getting the service they expect. Those are the dashboards. When we start to combine all that data, we create those centralized universal first panes of glass, that tell us something much richer than just the monitoring elements and give us the ability to start to automate solutions that bring in correlated data from all over that architecture.

Robb Boyd:         Are there certain industries or verticals or, I don't know, any other kind of subsection, I'm not thinking of at the moment, that lend themselves better to doing this? In other words, they get more bang for their buck getting on this early. I always think sometimes there's... Certainly we've seen a difference in companies, we've talked about this a lot in the COVID era here, as we talk about some companies that were kind of born... What's the right terminology? You can help me out a here. Remote first. They were kind of born already working in a more distributed fashion and thus are more comfortable with a reliance on things being elsewhere and monitoring tools that you have to rely on to stay on top of that. What should customers be looking for in terms of if they're watching this and they're going, "Oh, I would be a prime candidate for this." What kind of things do they normally see? Are there key indicators?

Ben:                 Yeah, I would say right off the bat, the more complicated and the more robust your application is, the more bang for the buck you're going to get. Especially when you look... I think it was before... We had a client back in the day who had a single transaction. We're not even talking a whole application, a single transaction that had over a hundred different elements that were a part of that. IT elements. That was before containerization came on. This was through VMs and stuff. And so now you're talking single transactions taking on thousands of hops and applications having hundreds of thousands, especially if you count endpoints and you're talking about edge computing.

                        That gets numerous and just the manpower to be able to keep all that up and running. If you can use AIOps to simplify that, as Tanner was saying, and put it in a single dashboard that enriches everything you're seeing, the other tools, you can deep dive into those once you've identified that problem and you know what you need to fix or to prevent, as we're monitoring this stuff. But that's really what it's about. As your applications keep getting more and more complex and you get more and more things in, that's where you're really going to get the bang for your buck.

Robb Boyd:         Well, you bring up something that I want to verify is it feels like, when you talk about that customer having one transaction that so has so many dependencies, if I'm saying that correctly, do you think most customers know what all the dependencies are on the most critical transactions? Is that really what I'm getting into here?

Ben:                 No. No. No chance.

Tanner:            No way.

Ben:                 I think the latest statistics, like the average IT job, it's three years before you change companies. And so if you think of not only the human turnover, but now you've got tools out there... There's new thing out, a harness that we work with at Cisco a lot, which has to do with development. It allows you to roll out code going from three or four roll-outs a quarter, to doing hundreds of roll-outs to thousands of roll-outs a quarter, just constant integration. And when you have that many developers constantly updating an application and constantly rolling out new code, it's impossible to keep up with all the changes that are going on.

Robb Boyd:         Yeah. And it's impossible for a human to keep up with it. Because I think that's key to what you're saying here too, is that you're building something that is not about increasing anyone's personal workload. If anything, it's about decreasing physical workload from a human perspective, because there's a lot of things that we just shouldn't be having to do. So, that's interesting. I like the fact that the dependency and just the fact that highlighting what these things can do for you. It feels like, in a lot of situations, that if any customer begins thinking about the fact that, if they don't know what their risk is for any given transaction, they may not even be able to actually prioritize what they should be most concerned about, investing in, or working on without this level of visibility. And this kind of thing really implies the entire infrastructure because you're looking at the stuff, people are going to feel. Yeah, Tanner?

Tanner:            Yeah. I was just saying, you just landed on something that's really important. And that is even, a lot of times we talk about AIOps or these big solutions as like they fixed some big problem. You may not have a problem. You may just have the inability to narrow down what's going on enough to actually focus on it. Nobody wants these war rooms. You're talking about very advanced tech companies that are still sending people on a call. They're on a giant WebEx, having 50 people trying to find the root cause. It's not because they're somehow technically incapable, it's because there are so many dependencies that are actively changing so often.

                        If you weren't doing something that is leveraging machine learning or AI, there's no way that you'll ever keep up with it much. Just in Silo, not talking about the whole enterprise.

Arsalan:            Right. And it's not just that it's to incorporate your legacy systems here, right? The older the company, the bigger the company, they have legacy systems. There's nobody to check those. It's next to impossible to check those.

Robb Boyd:         Well, Arsalan, now I'm going to put this on you then as we wind up here toward the end, I know it could be probably just as simple as a customer coming to World Wide Technology virtually or physically and saying, I want this AIOps thing, what next? But I wonder if you could maybe put a little bit more on that and say, what should someone be coming to you with? because I assume you're not saying that legacy applications disqualify anybody from doing this, [crosstalk] you're just saying it's a recognition of a different set of things that need to be considered and taken into consideration.

Arsalan:            Right.

Robb Boyd:         Okay. Well, you've all talked a little bit about what you guys have built towards. And I think this is the part I want to make sure it's not left behind, but as we end on this, you guys have been working together to say, "Let's build something, so we can not only build the tool set that says here's all the different things that we've learned through our own trial and error that we can demonstrate for you. And then we can work with you to apply it to your unique situations." But, Arsalan, this is something someone could come to you perhaps and learn from you on this and begin saying how this could... What are some of the resources that that WWT makes available, that you recommend people take advantage of in this area?

Arsalan:            One of the biggest things in our professional service is we incorporate the expertise of people that are developers that implement really complex systems, right? So it's a pretty critical thing. That's a pretty critical thing for our business itself. The one thing I look at for any customer that just want to come in with their own tool sets... Go ahead, provide a list. We'll take a look at what can be integrated what cannot be and we'll figure out a solution that works for them as well. And then on top of that, bring in our own solution on what's worked for us as well. So it's also about that. We're still learning. So it's at an advanced stage, but we're still adaptive. We're still adapting to their systems as well.

Tanner:            I'm going to add to that if I can, Rob.

Robb Boyd:         Please do.

Tanner:            Yeah. Go to the platform. The platform is key. This is where all of our work we do inside the ATC is really projected out onto the platform. So as we are progressing through integrations, we're progressing through architectural models. What we've learned, some of the expertise that you've heard us talk through is in articles and reports and videos and virtual labs and labs on demand. Things that you can... Or labs on request that you can go out and request a demo of the AIOps architecture and the goal is to help you along your journey. World Wide's whole focus here is to help people move down that journey. Go to the platform, register on it, get involved in it and start being part of the conversation.

Robb Boyd:         Well, it's one thing I love about what you guys have done with ATC is that I, and of course, I wonder what it smells like now with all these kitchen references, but I know it's a beautiful lab setting that I had the-

Tanner:            I don't think it's as good as it sounds.

Robb Boyd:         Yeah. No, no, no.

Tanner:            Smells like plastic and new stuff.

Robb Boyd:         Exactly. Constantly. There's a flow of stuff going on, because you guys are always doing so many proof of concepts. Not only that you've done yourselves, but I know you work very deeply with a lot of different vendors, because coming from a former vendor, I know we depended on you to be able to give us some additional technical reality that we couldn't get from customers. And it be very difficult for us to test ourselves. And so we needed someone like what you guys were doing. And I know that a lot of manufacturers depend on that. Final point on from AppD Ben, I'm just curious? Any final words on first steps? Feel free to be personal about this one and what you represent. But I know you've done a lot of work in this area and I appreciate your time being here.

Ben:                 Yeah. I'd say first steps, definitely reach out to WWT. What they bring with ATC, and I know we've touched on this, is the fact that you can go to individual vendors, but when it comes to putting everything together, nobody out there is equipped to do it like WWT is right now, with their ATC and what they have. And don't think you're going to come over to WWT, Tanner will probably hate me for saying this, don't think you're going to come over and all of a sudden you're going to get like, "Here's your AIOps. Now you owe us $500 million.

                        It is a journey. And it's something where we map it out. And I know between AppD and WWT, we do business value consulting to say, "Here you put these pieces in. Here's what you can expect to see out of this. Here's your return on investment. And you start little, you see those returns on investments and then you start really building it out. And it becomes that true journey, instead of just saying, you drop it in, you walk away. It is a journey. So reach out to WWT. They're uniquely positioned. We're happy to partner with them. I think we do a strong relationship in the marketplace and big things to come.

Robb Boyd:         Well, thank you. I appreciate your time, Ben. Arsalan, I appreciate yours. Tanner, I'll end with you and just ask you, we've got one more part to this series. What should we expect in part three?

Tanner:            Okay. So we've done all the talking we're going to do here. I think. Part three is really going to be about us digging into... Yeah, Rob's like, "Hey man, that's my job." We're digging into the architecture. So we really wanted to, instead of leading out with this... Ben said something there, that this is a journey. It takes a lot of pieces. It takes a lot of steps and you don't just get it. You can't just put your quarters in and get your AIOps. It involves a lot of about who the customer is. What their important things are? Their business objectives.

                        What we have done is we've created the very best demystified, generalized version of that architecture to help customers understand what's out there and how it works. Really, to take all the complexity out of this, take all the magic out of it and say, "You start with this and you move to this. You add these three components, you're able to read these things." And so the next chapter for us is really to say, "This is how we did it." Build our credibility. Help other people understand how we got here and then to really show them what we've created.

Robb Boyd:         Perfect. I encourage everybody of course, to subscribe to this whole series. This is not the only topic we talked. This just happens to be part two of a three part series. I'll tell you what I learned from you guys today that I really like is one, I feel like we stress the journey and the fact that, and I think anyone who has any level of self awareness in technology never would question that, right? That's what this is always about. But the fact is you guys have really worked hard in this journey already. And it feels like, from what I've learned from you, is that even with a small amount of time and investment, suddenly there's actually some early payoffs that for... I know a lot of customers who sometimes need to, "Hey, I need some quick wins maybe to justify further time and investment."

                        And it sounds like there's this an area that's ripe for being able to justify that, if you're an organization that maybe need something like that. It's amazing what a little bit of visibility can give you, but as you get started on that journey, the next thing that I really like about this is the fact that you guys have been there and you can share from multiple customers and help keep us from making mistakes that we might if you were just trudging through this and tripping through the forest on our own. I think that becomes really critical.

                        And I love the fact that the ATC has so much investment in, and if we didn't say it already, the platform as Tanner referred to it, is wwt.com. It's that simple. And you go in there and there is a ton of resources and you don't have to physically go to St. Louis or any other locations around the world. There's a lot of virtual interaction. It can take you quite far and is actually something you should be doing on a regular basis anyway.

                        So, regardless. Guys, thank you so much. I'll close us out here. You've been watching TEC37. This is the Technology Education and Collaboration podcast brought to you by World Wide Technologies, my name... World Wide Technology. My name is Rob Boyd. Thank you so much for joining us today. I hope you learned as much as I did. I look forward to seeing you on part three-