Cloud Success is about more than Technology (Guest: Mike Kavis)

Media Thumbnail
  • 0.5
  • 1
  • 1.25
  • 1.5
  • 1.75
  • 2
This is a podcast episode titled, Cloud Success is about more than Technology (Guest: Mike Kavis). The summary for this episode is: Over his decades of experience, Mike Kavis from Deloitte, has seen a lot of successful, and successful cloud implementations. One clear lesson - technology is easy, but people are hard. Mike discusses how innovation has to be more than just implementing technology and tools. Successful enterprises have to break down the old silos and rethink their culture, rather than merely adopting technology layered over legacy processes and approaches to collaboration.
What makes the transition so difficult?
01:16 MIN
An example of how COVID-19 has influenced a company's transition to the cloud
00:58 MIN
From generators to electrical outlets; From data servers to the cloud
01:50 MIN
Leave the eventual architecture at the trailer park and start with a clean slate on the cloud
01:18 MIN
Existing companies can adapt their culture and make it their own
01:33 MIN
A glimpse into AIOps
01:47 MIN

Mike Kavis: It's the same thing with the cloud. They just need that event to make people just forget about old ways of thinking, forget about control and let's advance the company and let's move it in this new direction.

Ben Newton: Welcome to the Masters of Data Podcast, the podcast that brings the human to data. And I'm your host, Ben Newton. Welcome everybody to another episode of the Masters of Data Podcast. And I'm, I'm excited about our guest today. Mike Kavis is a long time veteran of the cloud space. He's actually the Chief Cloud Architect at Deloitte's Cloud Practice and excited to talk to him today. Welcome, Mike, to the show.

Mike Kavis: Hey, thanks for having me always love talking cloud.

Ben Newton: Particularly in this situation, I think this has come up a lot, so I'm really excited to get into this and you really definitely have a really interesting background in talking about that. How did you end up where you're at? You have decades of experience in the cloud in particular, but what's kind of your story? Where'd you come from?

Mike Kavis: Yeah, so I worked a lot of small mid- sized companies my whole career and I was at this one company, which was doing targeted marketing and coupons at the point of sale. I was there for a very long time and every year they would say... We had dial ups, we had to dial into 30, 000 grocery stores and update the database. And every year they'd say," What's it going to take to do real time?" We'd do all this analysis and it was always more than our entire IT budget. So you get to about 2007 and I'm like," Hey, the cloud is the way to go." Instead of pushing everything down, the systems can just make a call up to the cloud and people are like," Yeah, you're crazy." So I left and we did a startup, amazingly at the same time, some guy came up with an idea similar to that and needed a technologist. And we got connected and I left and we basically did a digital coupon from Amazon really, so you go shopping and you buy your stuff in the cashier hit total and everything would up to the cloud and figure out what your coupons and rebates and all that stuff was. It would send it back down in milliseconds and it would come off your order. So, the old model was we had to put a server in every store. The new model is you didn't have to do anything, you just turn on some functionality and a point of sale. So that's how I got into it. That was like 2008 and did that for a few years. We actually won the Amazon challenge 2010, sold in 2013 or something like that, somewhere around there and I worked for that company for a while and basically said," I just want to help people figure out this cloud thing." So I've been consulting, Ooh, I don't know, five, six, seven years now. I was at Cloud Technology Partners for about four or five years and they got acquired and then I moved over to Deloitte where I've been a little over two years here. So I've seen a lot. I've seen it from when cloud was... When I got on Amazon, there was about eight APIs or storage disc queuing, that was it. And we were managing MySQL across multiple regions and zones ourselves. And then this thing like RDS came out and I'm like," Hmm. I don't have to do any of that work anymore." So it's been... Now you can pretty much do anything as a service. It's been a fascinating journey.

Ben Newton: Yeah. That's a good reminder because I think you see a lot of people that got into this recently and they're like," Well it's always been like that." And it's like," No, actually it was a storage in the beginning." It's amazing to see how fast it's changed though.

Mike Kavis: Yeah. I just remember we were going nuts because you can get a server for 50 cents and now that same server is like pennies.

Ben Newton: Yeah, yeah. That's true. Well, and I'm sure you have a somewhat similar background to that. And when I started out, I remember carrying around the 20, 30, 100 thousand dollars servers and I actually did some of the racking myself for a project and moving from that to where I probably, I didn't visit a data center for probably a decade. I hadn't seen the blinking lights.

Mike Kavis: Yeah. And where it's gone to, I mean, my daughter is in her early twenties, is certified on like Wix and all these platforms. And she builds websites for people. And someone asked her some infrastructure question and she texts me and says," What's a server?" And it kind of blew my mind, but it blew my mind because she's never had to deal with the infrastructure and she's making top quality websites for people because that's all been abstracted. And sometimes we just need to get out of hugging our old technology and think about that. Here's a 21 year old with a marketing degree that's building amazing websites and doesn't even know what a server is.

Ben Newton: Yeah, no, no, you're absolutely right because I remember when I was in college, building really ugly websites as a technologist. It's like," Hey, I can make texts blink. I'm so good at html." Well, not, that's really cool. And I think because you've definitely had a really interesting vantage point to see how companies have been adopting and how it's been changing things. And, one thing when you and I were talking before and a subject that I'm really passionate about is, I think it's changed, but you know, in general, when, cloud adoption really started accelerating, I think there was this idea as like," Oh, it's servers in the cloud, I'll go do that." And companies weren't really thinking that," Hey," to really take advantage, Netflix is not just servers in the cloud, they operate a different way or a lot of these...

Mike Kavis: Right.

Ben Newton: Cloud native companies, they operate a different way and there hasn't been a full recognition of that. So, I mean, what do you see from your side? I mean, you work with people every day that are trying to make that transition, right?

Mike Kavis: Yeah. I agree. And I'll just tell a little journey of what I've seen over the years. When I started in consulting, which was probably about 2013, it was all about, what's the ROI? I mean, people weren't even going to the cloud, they're trying to understand ROI. And a year later, everyone's starting to go there and it's like," How do I build on the cloud? How do I build?" So it was all about tech. And then around 2015 is like, what's this DevOps?" And everyone wanted to know about DevOps and new ways of working and then we got to like now, and it's like," Okay, I'm on the cloud. I've got CI/CD. How do I run this stuff? Everything's breaking." We went from a data center with defined processes, good or bad. If something broke, we kind of knew how to respond. Now we're in the cloud. And all we thought about was the tech and something breaks and it's like, what happens now? So now there's this big movement on, what's the right operating model? How do we need to organize? How do we rethink our processes? Before we go automate, let's fix some of these legacy steps. The step that got put in when we were deploying twice a year doesn't make sense when I'm trying to deploy twice a day. So, it's like the last two years, I've been more of a psychologist than a technologist. And I haven't really been doing much of the technology. It's all been about, how do we rethink organizational models? How do we shift roles and responsibilities closer to the people who know the product and service they're building? So that's where a lot of my work is now and I think that's where companies have a problem adopting cloud. They usually start with a group of really intelligent people who build something fantastic and have success, but they have a hard time scaling that to the rest of the company because it takes the whole company to change. It's more than five, six people figuring some stuff out.

Ben Newton: Yeah and no, that's really interesting. In particular, you remind me of when the DevOps discussions started. Call it, late 2000s, 2010 timeframe where they were really getting going. What was it all about? It was all about puppet versus chef. It's like, how do I... crosstalk Take you to... Yeah, yeah, yeah, for a lot of people, but it's really funny I remember we hosted a session with Gene Kim and he gave a presentation and it was all about organizational models and Taylorism versus... I mean, it was a little mind boggling. People were just eating it up, but it is really interesting that there's these cultural models and in particular, when you talk about making that transition, what are the things that companies are really struggling with? Because I mean, there's obviously a lot of ways to talk about, but what is it about the organizational models you think that are holding them back?

Mike Kavis: Change. Change is hard. So it's easy... I always say this, technology's easy, people are hard. You can always... Every place I go, regardless of how screwed up or how advanced they are, there's always brilliant people at these organizations. And they can always figure out the technology problems. But when it comes time to rethink organizational structures, rethink business processes, there's people's jobs that are tied to those things. There's people who've been trained to think a certain way for many years. It's just hard to break that mold and even when you get there, there's a VP who owns these 60 people that doesn't want to release it to... There's politics. So that stuff is so hard. And, something like COVID- 19, I guarantee you there're tons of companies that have been trying to go remote forever and ever and people just said," No, no, no." And then an event happened and they had to do it. You know what, they figured it out. It's the same thing with the cloud. They just need that event to make people just forget about old ways of thinking, forget about control and let's advance the company and let's move it in this new direction.

Ben Newton: You know, it's interesting the way you phrase that. So the event that drives change and I mean, I guess that's, as a psychologist, there's probably some psychological term for that, but what do you think that what's going on right now is that event for a lot of companies or is it overblown the idea that this is really driving change?

Mike Kavis: For some. So my colleague, Dave Linthicum, wrote an interesting article about this a few weeks ago, but someone he knows is a CIO all on- prem, and they have a major system that went down right after we quarantined everyone. And they needed to replace a lot of infrastructure and all the replacement infrastructure was shrink wrapped on the floor, ready to go, but nobody could go in. Everyone was quarantined.

Ben Newton: Oh yeah.

Mike Kavis: So it took a while. The system was down for awhile. And that particular CIO is like," Okay, I think it's time to rethink our fears of the cloud and security of having it and doing it ourselves. I mean, that's crazy. We can't let this happen again." So that particular company is, when we come out of this, they're going to be going full blown move towards the cloud and try to get out of situations like that. And that's just one story, I'm sure there's many, many others.

Ben Newton: Yeah, yeah, no, I've definitely heard them from my side too because even when you're talking, managing, the infrastructure, Amazon and Google and Facebook and all these other companies are running, Microsoft, that are running really large data center operations. I mean, they've already figured out how to do a lot of that stuff in a fairly automated way, in a way that would actually be compatible with social distancing. So they're actually, I mean for the most part, probably okay. Whereas if you're kind of even an old style data center operations, they're not equipped for that.

Mike Kavis: Well, that's your core competency. That's what they're in business to do. If I'm a healthcare provider, I'm not there to be great at infrastructure, I'm there to be great at healthcare. So we all know the story of electricity, right?

Ben Newton: Right.

Mike Kavis: But one story we probably don't know is that before we had outlets in the wall, the companies that could afford it would buy and manage their own generators.

Ben Newton: Mm- hmm( affirmative).

Mike Kavis: So they would buy these big machines and they would run it and they'd have a VP of electricity and a bunch of people underneath that person and that was one of the most critical things in the company because if the electricity went down, the machines were down, the assembly line was down, business was done. Right?

Ben Newton: Mm- hmm(affirmative).

Mike Kavis: And then all of a sudden, electricity was a utility that you plugged into the wall. So if you liken that to the cloud, you build and rack and stack machines, everything is in your data center and now it's available as utility. Why should I take the operating model from when I had to do this all myself to the cloud and apply it? Just as why, and I'm sure this happened, obviously I wasn't around in the fifties, but I'm sure the VP of electricity put up a big fight when we started plugging things into wall. We didn't need them anymore. And I'm sure they had checklists, lots of checklists because if power went down, they're doomed. I'm sure when they moved to an outlet in the wall, they probably brought a lot of those legacy check gates and stuff like that with them. It's the same thing that's happening in cloud. Now the difference is, over time, everyone's on an outlet and we don't do generators anymore. I don't think that's the case with a cloud, we'll always live in this hybrid environment, but we shouldn't hinder the adoption in the cloud to the processes and controls and everything that was putting in place when we were deploying twice a year, when we were buying, racking and stacking. We still need to do those things, but let's not constrain... What happens is we wind up using the cloud as an infrastructure, as a service only. So we commoditize the cloud and that's not really the value. That's valuable if you're eliminating data centers, but it's not valuable for building modern applications. The vantage in the cloud is that agility, moving up stack, using things like Amazon's RDS and managed service or GCPs BigQuery. I mean, you can't get that anywhere else. You know, that's the value in the cloud and too often we're taking our" best" practices and I always put best in quotes and moving it to the cloud and to tools and everything and the thinking and we wind up with a very expensive data center in the cloud.

Ben Newton: Yeah, yeah, no, no, it makes it... And it actually is a really interesting analogy because in particular, I mean, if you think about the development of the electrical grid and how that's changed over time, it's in some sense, some part of that has come back. I guess the hybrid version of that is that yeah, I have a backup generator and now you have people putting solar panels on buildings to balance energy needs, whatever. But in that sense, there's always this back and forth about where that's being managed, but at the end of the day, the companies are still not managing it as their core competency. It's not like you need a VP of electricity. You may decide to have that backup generator, but you're adopting that in a way that is like... You're not being tied to the past and I think that's pretty cool. Cool analogy.

Mike Kavis: Yeah, and the other point there is that it's not like we don't need someone thinking about electricity anymore. So when we plug into a wall, well now we still need to think about redundancy. So now we have multiple providers and we may play word games and what happens if this provider doesn't go? So it doesn't go away, it's the roles and responsibilities change. And that's why I talk so much about the operating model. You know, when we build in the cloud, like I said earlier, most places you throw it over the wall to a team that has all these processes. They may have a knock and they just, no matter how good or bad it is, they know what to do. Then you move to the cloud and you have nothing, right? It's Greenfield. You have nothing, right? And you could just do the same thing you did. Or you could... I like to say we have this eventual architecture, we have 30 years of spaghetti and that's our architecture today. You can put all that crap in the cloud and you can get a real spaghetti architecture because you can make change real good or you can take this once in a lifetime opportunity to start with a clean slate and build a new operating model, build a whole new platform to build and run software on and all that. And that's the mistake a lot of people make. They just bring everything forward. In 30 years of trailer park architecture, I have this one slide where there's all these trailer parks all over the place on stilts with taped up windows and eventual architecture. No one designs it that way, it just happened. Over the years you stick stuff together. Don't bring that to the cloud. This is your opportunity to do something right. Get the foundations right.

Ben Newton: I think I really liked that trailer park architecture. It reminds me of... I saw somebody else that I've used a picture in a few presentations where they had, what you see in Brazil and some of those neighborhoods where people are basically stealing power and there's like cords go...

Mike Kavis: Oh yeah.

Ben Newton: There's literally like a massive cave and they called it like building... What was it? Skyscraper Favelas in code or something like that. But basically, you're taking that same kind of thing and building it in the cloud. But I think the trailer park architecture is a little bit more accessible. I like that.

Mike Kavis: Well, the funny thing about that slide is I created a presentation called SOA and Change. Remember when SOA was a thing?

Ben Newton: Yeah, yeah.

Mike Kavis: And that slide was in there and that deck was, I think, 2006. I can do a replace all SOA with cloud and that deck would work perfectly because it's the same thing. When we adopted move from the mainframe to windows and client server, when we adopted the internet, when we started doing an ERP, it's change and organizational change management, operating model changes, those are the things we fail at. We always figure out the tech, but we make the journey so much harder and more expensive than it has to be because we only focus on tech.

Ben Newton: Yeah. Yeah, absolutely and one thing I'd love to ask you about because it's something that I've been really interested in myself and I'm always interested to see what people are saying because one of my favorite books on the subject by Stephen O'Grady over at RedMonk, he wrote on The New Kingmakers, but he makes this argument that basically cloud unleashed the software engineering organizations because IT wasn't the gatekeeper to compute anymore. They weren't the gatekeeper to the architecture where you could go build things and when they actually could literally go and swipe a credit card, then suddenly it started changing their operating model, so you started seeing the power shift to these engineering organizations and if you get in in a wider view... It seems like there's kind of a, when you talk about that people architecture shift, that organizational kind of change, it seems like it's more than just," Okay, we're going to have a DevOps team in our IT organization. There's kind of a fundamental ownership operating model and new relationship with the people who actually build the code. I mean, what do you see?

Mike Kavis: I agree with that. So, before I started my startup journey, we were on this one project where we wanted to buy like 10, 1U blade servers and we had to go through this department who had a different set of priorities than us, and honestly didn't even like us. And eight months later, I'm still waiting for that and then like a year later, they're finally given it to me and I'm like," Well, these are outdated. I want newer ones." And then I left for a startup. I went right click 10 servers, bam. And I'm like," I got some power now. I can do some things." And the other thing about that is a cloud without those handcuffs to have to go to this person for infrastructure, I can figure out what the best infrastructure is for me. So again, going back to that company, regardless of what my business problem was, I had to use this big IBM box. And that was shared with everyone else, regardless of what my problem was. And regardless of what my problem was, I had to use this Oracle database. And then I go to the cloud and I'm like," Well, I need a graph database for this. I need ElastiCache for this. I need a relational database for this and I need a dupe for that. Same app, right?

Ben Newton: Yeah.

Mike Kavis: Or same collection of apps. And you know what? I don't have to hire Hadoop people. I don't have to hire graph database people. I don't have to... I can just do this myself because it's abstracted. Same thing with my daughter, what's a server? I'm not great at setting up ElastiCache but now I don't have to. So we've kind of freed developers to experiment and use the best technology instead of being forced to use what we have golf shirts for. That's what we all used to joke," Oh, I got this golf shirt and it says, Netezza therefore everything must be Netezza. Really? Not really. Just because you had lunch with them 38 times doesn't mean I have to use that. Right? I want to use what meets my business problem and the application I'm working on tomorrow has totally different requirements and I may use Mongo on that because I can now. So yeah, it empowers them. The danger is the developers sometimes think they don't need these people. And then the developers often aren't experts at security, aren't experts at operations, aren't experts at networks. So you still need them. You need their engineering, their architecture insights. You don't need their solution so much anymore. You need their minds. And that's the difference. We used to need their products. Now we need their expertise. And that's why we talk about shift left and full stack teams and all that. We still need these people, but we need to work together to solve solutions and we need to have the same shared goals so we're all working towards getting a customer what they need. Instead, what I was used to is, I used to love this, we go into this meeting and I had my top 10 priorities and the database team had their top 10 and the network team had their top 10 and none of mine were on theirs. It's like, how does this even work? It doesn't.

Ben Newton: Yeah. No, you're absolutely right and I remember, having come from that exact same background and I remember having a conversation with, let's say a millennial software engineer, and basically talking to him and I was like... And he's talking about setting up a database and doing all this work, and then I realized that it's the logic. We don't have a DBA and I remember to your point, one of my first really big consulting gigs back in the day, is like Oracle was like the center of the world, everything revolved around this ginormous Oracle cluster that we build out and the DBAs kind of ruled the day. And I think now to your point, it's not about, at least for a lot of what I see seen, it's not about the A in the DBA, it's not about the administration. It's about the architecture. Well, what's the right way to set this stuff up. How should you think about it? Okay, so you've used the automated version of X, Y, and Z, but you know, now things are running really slowly because you don't really understand how that kind of structure works. It shifts the conversation from, I spent my whole life just managing this Oracle cluster and you tell me what you need and if it fits in my boxes, I'll deliver it for you. One thing I want to ask for you too, in that context is like what you've seen because that was one of the things I thought was interesting about Stephen O'Grady's book, and in particular, I've spent a lot of time recently for various different reasons kind of looking into like the SRE model, the site reliability engineering model, and how it's different, how it isn't different and how it fits in and how people are thinking about it. And at least for me, like one thing that it kind of boiled down to was, in some sense, who ends up taking the operational responsibilities? For me, DevOps was a movement that like you said," shift left." It was, okay so the IT organization is going to be proactive and they're going to learn enough about what the software engineers are doing so that they can collaborate and they can work together to deliver a better solution. Whereas it seemed to me to some degree, site reliability engineering was, you know what, if just start from just engineers and we forget that there ever were operational people in the world, this is how we would do it. Do you see any of that? How are these more traditional companies working with that? Because of course, if you're just a startup and you're starting from scratch, it isn't even a question, you just hire smart people and you do stuff. But like, to your point, when you have an existing company, there's existing culture and you can't just change that. I mean, how do you see that back and forth working out?

Mike Kavis: It's interesting. I'll crack a joke about it first. You know, first we wrote a Chef script and called ourselves a DevOps engineer, it didn't change anything. Then we renamed ourself SREs and didn't change anything. That's what a lot of companies are doing. And then there're other companies who get the whole Google model of, you build it, you run it, let's pair an SRE team with the Dev team, all that. The problem is, they're trying to be Google and they're not. So the first thing they do is they set up a center of excellence for SRE, so it's over before we start. And then they try to serve SRE from a central place. And the Dev teams are like," I don't need you." Because they're still trying to do it in a silo. And this comes to becoming a product centric company, or at least within a business unit, you're building something. You have to think product. And when you think product and the person with the budget owns the Dev and the Ops and the Ops could be the SRE model, then it kind of works. But it's like Spotify. Everyone wants to be the Spotify model and when Spotify first publishes your stuff, they basically said, don't copy this. This is where we are today, tomorrow we're going to be someplace else. And everyone's copying it. It's the same thing with SRE. Oh, let's do exactly what Google says. No, look at the problem they tried to solve, look at how they solved it. Look at their principles. Those are good principles. Take those and make it yours. Change some things, do what works within your culture. It's like day trading. We're just going to hit the home run today. The same thing with technology," Oh, Google did this, Facebook did this, so we're going to do it." It's like, you're not Google, you're not Facebook. But look at the problem statement, look how they solved it, learn from that and adjust.

Ben Newton: Yeah. I like the way you say product centric and I guess one part of that, how are you seeing that balance kind of shifting between the product business unit in a CIO model? Because at least the way I always viewed it, the CIOs were originally really kind of created because everybody was running a server in their closet. And every business unit in the office had their own set of servers in the CIOs kind of existed as a role to bring it in and kind of create a common shared platform. Do you find that that is what is happening? A lot of these companies that are doing it successfully is that the power is shifting to the BU or the CIO is doing something differently? I mean, what do you see?

Mike Kavis: I think that's where it's heading. I don't know if you've ever been to any of these DevOps Enterprise Summits, those are the best. And one of the things they do that I like is they have kind of repeat, I call them repeat offenders, but repeat presenters. So every year you get to see the progression of their story. So the folks who've been at it for a long time, they're definitely heading down that path. They're shifting everything left. They're going to T- shaped teams, t- shaped meaning they have all the skills they need and one person's managing both Dev and Ops. Now they still have core teams, core platform teams, core security teams and all that, but the teams that are building product or product focused and they have a full stack team and they have everything they need. They may lean on those centers of excellence over there or whatever, but they have what they need. That's where things are heading. I read an interesting article on LinkedIn a couple months ago. It was a major financial company, I think. And the thing was that they eliminated the CIO position. They had a CIO and CTO and they eliminated it and just had a CTO position. And it was very interesting. Basically saying," We're a technology company. We need to focus on technology." I haven't seen it that wide spread, I just saw this one company, but it was a big financial company. I was kind of blown away from that. I think, you go back to the electricity thing. As soon as plugs were in the wall, that didn't happen overnight. Same thing, they made these investments and all these things that required the generator. It happened over many, many years and that's the same thing that's happening here. It's just the people who've been at it for longer are further along what we're describing here as pushing stuff out to the business units.

Ben Newton: Yeah, the CIO to CTO thing, I can definitely validate. I've actually heard that. I mean, it's still a subset, but it seems to be happening more and I've seen stories about CIOs kind of repositioning themselves. But yeah, it seems like when companies that really want to make change, sometimes they put in the CTO and I've definitely seen that a nontrivial amount of times now where there is something going on there or like things shifting around. And again, I think one of the reasons why I find that really interesting is because early on in my career in the early 2000s is when the CIO position was becoming the big thing and to see how that's all kind of turned around. So I guess to kind of put a bow on all this, now that you're looking out, after the situation we're in, around COVID and things start to get back to some sort of new normal, I mean, what are you seeing? What are you keeping track of? What kind of rises to the top for you when you're watching?

Mike Kavis: Well, I'm a different beast. I'm always out on what's new. So, I was looking at this cloud thing 2006, 2007, right? I'm all over this AI, AIOps stuff. I'm really looking at that. And when I had my startup, I was kind of the Ops guy along with raising money and doing all these other things. I didn't have time to code all the time. I coded very little actually because I was on the road raising money, but I was living in New Relic and Nagios and all these things, I was kind of the Ops guy and out of necessity, we had to be very proactive. So I built this kind of model of... There were kind of phases of it. There's reactive Ops and there's kind of proactive Ops, which is the area I lived in where you have these tools to tell you when trends are going South. Then you get to like augmented. You get to intelligence Ops where it's augmented and it's fully autonomous, if you think about self driving cars. And there's use cases for all of that. So when you start thinking about IOT and stuff, you have thousands of sensors all over the place. That has to be pretty autonomous, you can't go out in the middle of cornfields all over the place and start messing with sensors if you're an agtech and stuff. So, I'm learning a lot in that space. When do you use the different models? What are the technologies that enable that? There're some really cool platforms now, of course, a lot of these were logging tools now they call themselves AI tools. But regardless, there's a lot of platforms for aggregating data, for helping you get to autonomous or fully automated and all that stuff. And I'm a sponge right there right now and I think that's a way forward. Because when you start really getting good at the cloud... We used to have three tier architectures. The place I worked at, all the servers were named after Star Wars, Battleship. Kind of the pet and cattle thing, Randy Bias always talks about and you're getting to this highly distributed model and servers come and go, humans don't scale the watch desk stuff, so you have to leverage intelligence. And the thing I always say is you have to trust in your automation. People don't trust, they still need their checklists and stuff. You can't scale that way. You have to trust in your automation, which means you have to monitor your automation. It's whole new area. Andy Mann runs this new panel called New Ops, which I've been to. Really, people are rethinking Ops.

Ben Newton: Yeah. Well, Mike, this was a great conversation. I appreciate you coming on. We'll have to have you back on another time to talk more about the AIOps stuff because I think we could spend a whole episode just talking about that.

Mike Kavis: I would love to. Maybe I'll know a lot more about it next time.

Ben Newton: Okay. Okay. Well, I'll ping you to see, you can give me the temperature. But I really appreciate you coming on and I think this was a great discussion. Thanks for your time, Mike. And thanks everybody for listening and as always find us and rate us in iTunes or your favorite podcast app, and look for the next episode in your feed. Thanks everybody.

Speaker 3: Masters of Data is brought to you by Sumo Logic. Sumo Logic is a cloud native machine data analytics platform delivering real time, continuous intelligence as a service, to build, run, and secure modern applications. Sumo Logic empowers the people who power modern business. For more information, go to Sumologic. com. For more on Masters of Data, go to mastersofdata. com and subscribe and spread the word by rating us on iTunes or your favorite podcast app.


In this epsiode of Masters of Data, we sat down with Mike Kavis, Chief Cloud Architect and Managing Director at Deloitte's Cloud Practice, and talked everything cloud. In recent years, we've seen more and more companies transition to the cloud leaving behind their data servers. The cloud provides game-changing services including quick access to databases, files, and applications and easy data recovery. Even with the seemingly unlimited potential that the cloud holds, why have companies delayed this transition from servers to the cloud? And even if some companies go through with the transition, why isn't it always successful? 

Through our conversation, we learn that one thing is for sure; figuring out the technology is the easy part, it's people that bring the complications. Change is difficult and especially during this pandemic, having to adjust is vital to the success of a company and its people. With the cloud in particular, change should be embraced with open arms as it provides companies a clean slate to create a new operating model for themselves.

To learn more about Mike Kavis and his work at Deloitte, connect with him on LinkedIn.*

And to learn more from the team, visit Deloitte's collection of technology-focused articles below.

*Links found in resources section below