Clay Magouryk - SVP of Oracle Cloud Infrastructure

(Joel Beasley at 00:00:00) Hello, my friends. Today we are talking to Clay, the SVP of Engineering at Oracle Cloud Infrastructure, and we discuss Oracle's move into the cloud infrastructure space, how responsibilities change as an organization matures, and attracting top talent with interesting problems and an audacious mission. All of this right here, right now on the Modern CTO Podcast. Here we go. This is the Modern CTO Podcast.

(Joel Beasley at 00:00:34) So how did you originally—give me the origin story of Clay. How did you fall in love with technology?

(Clay at 00:00:39) Well, growing up, my dad was a programmer. He was in the military. He was in the Air Force and did a bunch of programming stuff when he was there. And so when I was growing up as a kid, we had a computer at home, and so I would play DOS games and things like Doom and Civilization on a PC at home. And then when I was 10 or 11, I ended up getting into programming.

(Clay at 00:01:02) I don't remember exactly what started it. Part of it was the Internet kind of came about and I got the Internet, and I was like, okay, I need to make websites. So you go and you get a GeoCities account and you start learning how to make things with CGI scripts, and you learn Perl back in the Perl 4 days when they didn't even have nested data structures, and you just start doing things.

(Clay at 00:01:23) And I really liked programming. And so then I bought the Kernighan and Ritchie C book, and I worked my way through the entire C book. And then I just did a bunch of—for two years, from 11 to 13 or something, I just programmed all the time. And that was really how I got into it. I think it was just having access to computers, and it's interesting.

(Clay at 00:01:43) I've tried to help a lot of other people get into programming. I find that I'm not very good at it, because for me, I just share, oh, look, here it is. And they go, yeah. And I was like, well, then we're done. Now that you understand it, obviously you will just do this all the time, right? And they go, no, this is not fun. And I say, well, then you should just give up, because I don't know how to help you.

(Clay at 00:02:03) So not a very good teacher of—because for me, it was very simple. Once I realized it was possible, I was like, well, I'm gonna do this.

(Joel Beasley at 00:02:12) Do you ever get the parents who want their kids to do it because they see other people making money in it? I used to get calls like this all the time. Oh, will you come over and give Johnny an intro to one-on-one programming because you have an app company? And I'd be like, the first couple times, sure, this is great. Another person is interested in this. Then I realized, no, the kid was doing it for a Chili's gift card or something. No interest in it.

(Clay at 00:02:36) I struggle with that even with people in my family, nieces or nephews, where they're like, oh, they kind of seem interested in it. I buy them books and I talk to them about it, and then you wait a week and you're like, well, did you do that stuff? And they're like, nah. And I go, well, okay. I guess—because I'm not qualified to really have an opinion about how to educate people on programming. I have not done enough research, nor have I deeply invested in it. What I do know is that for some people, it's very visceral. They experience it and it's obvious. My hope is that other people can also like it. Maybe it just takes longer. But I do think there's a big chunk of people who end up getting into it because it is lucrative and plentiful in the job market.

(Clay at 00:03:18) And it's not for them, and then they end up burning out pretty quickly and switching jobs. So yeah, I have gotten a lot of those kinds of things. And it's funny, you would think along the way I would at least find one person that I could help. Never really kind of happened.

(Joel Beasley at 00:03:35) Yeah. I've took the perspective—somebody told me once the story about a monk, and you have to ask the monk 10 times for help before they'll help you. Right? And it's like, okay, I'm gonna adopt that philosophy because it's way too much work for me to pour energy into people who aren't working their face off trying. So now I've got this thing where I'll take people who are running and help point them in the right direction because they've got a lot of energy and momentum behind them, but I'm not gonna try to convince someone to run.

(Clay at 00:04:05) Yeah, I completely agree. And that was—it's funny because when I have conversations with people about this stuff, I feel like programming is so easy to get into. I don't think there's a lot of other industries where you have all the same tools as the best professionals. Right? For example, I also enjoy certain sports, and some of those sports are very expensive. It costs a lot of money to practice those things. Well, programming, if you've got any kind of a computer, you have the best tools that basically anybody has. And that's not entirely true these days. If you want to do certain machine learning things, you need access to GPUs, et cetera.

(Clay at 00:04:40) But still, for a couple thousand dollars in a home PC, you can do whatever you need to do. But then the only limitation becomes on yourself and your ability to actually sit down, do the work, read the material, and work hard. And I really enjoy that about programming.

(Joel Beasley at 00:04:55) You're right, because if I want to be a biologist, I don't have multimillion dollar microscopes and labs right around me. I can't do that. I literally can't. I don't have access to it.

(Clay at 00:05:06) Right. Or you can't travel to go to different places to see different plants or animals. Or almost any—I went to college for electrical engineering, and electrical engineering, you need all this stuff. And if you're a power distribution person, you have to have these complicated power distribution labs. Everyone has access to what they need to be a programmer, and yet the vast majority of people don't pick up on that.

(Joel Beasley at 00:05:27) But you do, and you have a team of over a thousand people that work with you over at Oracle that get it, right?

(Clay at 00:05:34) Absolutely. But I don't know how you get—I don't know what causes that. I don't know if it's built-in personality or if it's something that can be trained. I wonder sometimes if there's a magic secret formula where if you find people at the right stage in their life, maybe you introduce programming at the right time during school or something. But in general, I'm very glad that we can build successful companies.

(Clay at 00:05:59) And here at Oracle, we've hired more than 3,000 to 4,000 people that really love programming. They come to work every day, and they work very hard and build amazing things.

(Joel Beasley at 00:06:08) Yeah. I've noticed it's like gravity. You get enough cluster of the certain quality of people you want and they'll just—it'll attract like gravity. And so I focus more on that than questioning the why behind it. I'm like, alright, we just know it works. Let's just get these people together. And that seems to be what's happened over at Oracle. You guys attract some of the best talent in the world, right?

(Clay at 00:06:33) Yeah. I think—well, I think there's a few key things to attracting good people, right? I joined Oracle five and a half years ago, and there were four of us in a tiny little corner of an office building in Seattle. And now we have more than 5,000 people working on this overall effort we call Oracle Cloud Infrastructure.

(Clay at 00:06:50) And for us to get there, a lot of it came down to, well, how do you attract people? Because it turns out it's really hard to build a cloud with four people. And I got a really quick crash course on how do you get a team together. And having done hundreds of interviews and hired hundreds of people, it's really quite simple. People want to make sure—there's the basics, right? People want to have a place to work where there's an office or they can work remotely, and they want to make sure they get paid. And there's certain things that matter like that. But really what attracts people is who they work with, the people they work for, and the work they're doing and the mission that all comes together. And so if you have a very attractive mission—an ambitious goal that is not guaranteed to succeed but very exciting—and you have interesting problems to solve to get to that goal, and you have good coworkers and great people that are leaders, that's all you need.

(Clay at 00:07:44) And of course, you still have to pay them and make sure they have a place to sit and give them some computers and things like that. But really, it's about having those overall mission pieces together. And what I love about that is that there's no limitation on those things, right? Those are all something that as a company or as a manager or as a team, you get to decide what you should be doing. And you can either choose to be ambitious and interesting, or you can choose to be boring and myopic. I think in general, you'll be more successful by being more ambitious. It attracts better people, and they'll be much happier to work there as well.

(Joel Beasley at 00:08:20) Absolutely. Now you mentioned something—you said four people to thousands. Were you on the founding team of this division that you're a part of now?

(Clay at 00:08:28) Yeah. So there were when I joined, there was four of us. And that was five and a half years ago now. And so it's been a pretty crazy ride over the past five and a half years.

(Joel Beasley at 00:08:38) So did you have to learn P&L-type responsibilities? Or did you already have those skills? Or are they not necessary because you have other people on those four that deal with that?

(Clay at 00:08:47) I had to learn a lot. So my background when I joined Oracle—my background is I used to be a software developer. I was at Amazon for many years and I wrote software. And obviously, coming from Amazon, you have a lot of operational background. So you understand how to maintain services and do operations and that kind of stuff, and you know how to hire and build teams. But there was a whole lot of giant crash course over the past few years about understanding what does it take to actually run a large organization? What are the responsibilities from a business perspective? How do you even just communicate the financial perspective? And how do you manage things like massive amounts of capacity?

(Clay at 00:09:23) How do you build data centers all around the world? As you grow and scale a larger team, one of the things that's also interesting and difficult is you're no longer managing a team of people that are like you in the sense that when you have a small dev team, you typically manage a bunch of developers. Well, when you get to a large organization, you have all different types of people. You have finance people. You have operations people. You have designers. You have technical content writers. And so then how do you actually effectively manage all those different types of roles in a single organization, keep them motivated? Yeah.

(Clay at 00:09:54) There's—it's been a very interesting and fast ride over the past few years.

(Joel Beasley at 00:09:59) Yeah. But it's fun, though, right? Because you're always learning new things and expanding your experience.

(Clay at 00:10:06) I've never had an opportunity like this in my life before, and the opportunity I've been given at Oracle is something I'm thankful for every day. It's not always clear to me exactly why they gave me the opportunity. Sometimes you look back and go, well, I'm not sure that I was entirely qualified at each step of the way, but it seems to have worked out so far, and it's been extremely rewarding for me.

(Joel Beasley at 00:10:27) Yeah. Well, the opportunity floated in front of you and you had the ability to stand up and grab it, right?

(Clay at 00:10:34) Yeah. And I think that—just to be clear what I view the opportunity as—it's not just the opportunity to grow my personal career, right? In large part, I joined Oracle to make Oracle a successful cloud infrastructure provider. When I joined Oracle five and a half years ago, Oracle was already a very successful technology provider. It has some of the most successful technology pieces in the world, things like Oracle Database, but also MySQL, Java programming language, things like Oracle Linux, middleware pieces like WebLogic, et cetera. And then obviously, Oracle was already very successful in the application space. Previously, on-prem applications, things like E-Business Suite, JD Edwards, PeopleSoft, Siebel. And then more recently, but still very successfully, things like Fusion Applications, where—ERP systems, HCM systems, supply chain management, et cetera. So Oracle is already very successful in all of those areas.

(Clay at 00:11:25) And when I joined, the opportunity really was, well, let's make Oracle that successful, that relevant in the cloud platform space. And so, having worked at Amazon for many years, I knew how difficult that was going to be. But I also knew that this was the closest I could kind of get to doing a cloud infrastructure startup. There's not a lot of VCs willing to give you many, many billions of dollars a year to throw into a giant pit to one day actually have a successful worldwide global cloud-scale set of systems. And Oracle was both crazy enough and smart enough to know that that was what they needed to do and then hire some people and make it happen.

(Clay at 00:12:11) And so the opportunity I joined Oracle for is still the same thing that I'm here today working on. It's still the same thing that I sell candidates on all the time. This cloud infrastructure space is going to be very, very large. I think we're still very early on in the overall adoption of cloud. And there's going to be a relatively few number of players. It's not gonna be 10. It's not gonna be one. And when we make Oracle one of those players, it's going to be a very, very successful business. And also, we're going to solve a lot of the problems that our customers have. Our customers just aren't realizing the benefits of the cloud today, and we're gonna be able to deliver that value to them very, very soon.

(Joel Beasley at 00:12:49) So, alright, so I'm kind of learning on the fly here, because I did not know this. So wait. I, as an engineer, can go to Oracle and get a cloud instance? Like I go to Amazon?

(Clay at 00:13:04) Yeah. Of course. That's literally what my team does.

(Joel Beasley at 00:13:08) Well, see, look, I talk to people saying cloud infrastructure is great, but I don't know that it's cloud infrastructure like directly to consumer. A lot of these companies, these bigger companies, they're doing B2B stuff like, oh, my top 10 people were providing cloud services. You can talk to our sales engineering teams. Like, you could—it's like it's like competitor to Microsoft, Amazon Web Services. You could just go and get an instance.

(Clay at 00:13:31) Yeah. Of course. Yeah. I mean, look, go to oracle.com. Sign up for a free trial. We also have this thing we call the Always Free Tier. You get a few VMs and load balancers, some Oracle Autonomous Databases. Free forever. Build a little website. Look, I mean, that's what I came here to build. That's what we've built. We still have a long way to go. But we built new physical data centers, physical network. We have compute, block storage, file storage, object storage, virtual networking, load balancing services. You have Kubernetes service, functions.

(Clay at 00:14:05) So there's a huge amount of investment. That's exactly what we're doing.

(Joel Beasley at 00:14:08) Dude, if you go to the website, it looks like it's so simple now. It's literally just—it's a beautiful kind of—it's got this artistic feel to it, and it's just—I'm on just oracle.com, and it's like cloud, cloud, cloud. It's like Oracle's becoming this cloud computing company. That's amazing.

(Clay at 00:14:31) It's not just becoming. I mean, look, we've been working on this for many, many years. And, like, for example, the new website came out a couple months ago. It's a huge upgrade from where we were before. And we do absolutely have an awareness problem where the thing about Oracle, right, is that the vast majority of our customers are large, successful businesses.

(Clay at 00:14:49) And so we end up spending a lot of our time talking to them. But at the same time, we're also building a very general purpose, easy to use consumer product for individual end developers. And, you know, you go, you sign up with a credit card, you get an account, you get a 30-day free trial where you get $300 in credit. But once those credits expire, you can still continue using the free tier forever. It doesn't expire ever.

(Clay at 00:15:14) You can build, and it's a full feature, you know. You get a nice UI. Everything's available via API and SDK. You have a CLI. It's exactly what you expect.

(Joel Beasley at 00:15:23) Yeah. This is awesome. I had no idea. This is actually really cool. But you know what?

(Joel Beasley at 00:15:27) I watched this documentary on your founder, you know, Larry. And he's really big about being the best and being number one, which is great because I'm a huge fan of, you know, always putting your best foot forward and, you know, going through whatever pain you need to get to, to come out better on the other side. That's my deal. Right? And so to see you guys enter cloud in such a huge way is actually really cool.

(Clay at 00:15:50) It is. And that's the thing. You know, what I think makes entering the cloud infrastructure market so difficult is that it's such a huge investment. Right? Like, so as an example, you know, we launched what we call Oracle Cloud Infrastructure three years ago.

(Clay at 00:16:04) At the time, we had one region available in the U.S., and we had, like, five services. You know, as of now, we have 18 regions live. And over the next year, by the end of 2020, we'll have 36 regions globally. It takes a lot of work to go and build all those data centers and deploy those regions and operate them successfully. Same thing around, you know, compliance certification.

(Clay at 00:16:24) You have to go and you have to get PCI compliance and SOC compliance and have a HIPAA attestation. And then you have to go and get things like FedRAMP Moderate and FedRAMP High. And then you go and you build national security regions for the U.S. government and other countries' government. There's just so much work that goes into this business. And because the bar is so high, to really see that return on the investment takes a long time.

(Clay at 00:16:47) And that's what I really admire about Oracle, is the fact that it's a company that's willing to make that investment. You know, if you take a—if you go back a few years ago, there were so many companies that were trying to be in a cloud player, and they've all fallen off now, mostly because they just didn't have the intestinal fortitude to invest and continue investing and believe in themselves to build out the infrastructure and to hire the team and invest enough in it. And so what I really, like I said, super excited and admired about Oracle is the fact that it is a company that, while we entered cloud infrastructure late, totally willing to put the energy in to catch up and become a real player.

(Joel Beasley at 00:17:24) Yeah. That's a cool story, though. That's like a cool journey, especially to tell through—who else was, like, who are you really close with of the other four people that are pretty awesome?

(Clay at 00:17:35) Yeah. So, well, the guy who hired me, my boss, his name is Don Johnson. He still works here. He's the person who runs all of Oracle Cloud Infrastructure. I run the engineering portion, and he has that plus the overall business stuff.

(Clay at 00:17:50) And Don and I have worked together for a long time. You know, Don was an early AWS developer and worked there for a long time. And so we have a really good relationship. And then one of the people that was instrumental early on, his name was Craig Kelly. He kind of had more of the data center and networking, kind of physical bottoms-up view of the world.

(Clay at 00:18:09) And he's actually left us now. He retired from Oracle, I think last year. But he was with us for a few years and really helped build out, you know, the data center team and the physical networking team and some of the lower-level operational stuff. He was very, very good in that area.

(Joel Beasley at 00:18:25) No. Those are—Don Johnson, I love it because it's like a famous person's name too. It's very memorable. So, like, you handle all the engineering portion. I've actually been reading a pretty interesting book.

(Joel Beasley at 00:18:37) I keep interesting books on my desk. I also got the Elon Musk book. But this is newer. It came out a couple months ago. I came across it because the authors, David and Jessica, they were on my show.

(Joel Beasley at 00:18:47) But basically, I wanted—I was thinking about writing a book about product and, you know, my experience building product for the past 17 years and systems and scaling them. And then I read this one called Product Mindset about how to think about product, and it's a lot for, like, you know, executives, people like that, engineers. And it's just like a brilliant book. It's got everything I wanted. So what I did was, I bought, like, 10 copies of it, and then I said, okay, the next people that I feel the urge to share this with, I'll just send them a copy.

(Joel Beasley at 00:19:19) So I'm gonna mail you a physical copy of this. Is that cool?

(Clay at 00:19:22) Okay. Yeah. Is this the book that you wish you had written but didn't do it yet?

(Joel Beasley at 00:19:26) Oh, yeah. Well, so now I don't feel the pressure to write it anymore. Yeah.

(Clay at 00:19:30) Yeah. Now you don't have to.

(Joel Beasley at 00:19:31) Now I don't have to. Now I can just say, "Go read this," and people—it's already done. But, yeah, really cool company. But I always recommend it to people who, uh, senior engineering-type people, because you'll know, like, 80% of the stuff in there, but the way it's written, there will be—you'll walk away with, like, one or two really good ideas.

(Clay at 00:19:55) Yeah. Yeah. What would you say is the thing that kind of struck you the most about the book?

(Joel Beasley at 00:20:00) I would think that, like, the way that the teams interact with the customers—like, so the product mindset is a concept, like, how you think about product, making sure that it's generating revenue, obviously, something that's very important, like, self-sustaining. And then it's just got a lot of the newer sets of values about, like, communication and being very explicit about what value does this bring to the customer. Like, Amazon has their concept where, you know, they write the press release and things like that. But it just reinforces these concepts. It also talks about, like, selling your ideas to others in the executive level of the organization, like how you can actually push this product.

(Joel Beasley at 00:20:36) And it's just a well-written book. And I really liked the authors. And I actually—I didn't read the book. Like, I only skimmed it before they came on the show. And then they were just so bright.

(Joel Beasley at 00:20:48) I've seen the link to the episode. And they had so many good things talking about the book. I actually went and read it, the full thing, after the episode. And I think this was really good.

(Clay at 00:20:55) That's cool.

(Joel Beasley at 00:20:58) Yeah. So what do you value when building products, like, within your cloud?

(Clay at 00:21:04) So I think—well, I'll take us on a tangent and we can come back. I think one of the most difficult things about building a cloud is that you're not actually building a product, you're building a platform. And so, as an example, you know, when you show up and you use, you know, OCI, you don't use any individual service. Right? You have a compute service.

(Clay at 00:21:24) You have a virtual networking service. You have a load balancing service. And to be able to effectively deliver all of that functionality, internally, you have to break those things up into individual separate services. And then, obviously, those separate services have different teams. But when a customer uses those services together, when it uses the platform, they don't care about that.

(Clay at 00:21:46) Right? If what you want to do is you want to launch, you know, a website—well, you want to create a web application firewall. You want to put a load balancer behind that. You want to spin up a couple of VMs behind that. You might need a DNS entry. That spans a bunch of different services, but the customer doesn't care.

(Clay at 00:22:02) They just want their problem solved. Right? They want the solution to this issue. And so I think a large part of the difficulty of building this platform is about making sure that the puzzle pieces really fit together well and that you don't expose your internal organization or your, you know, service factoring to the customer. So I spend a lot of my time, and a lot of our product management time—yes, it's important to figure out what's right to build, but I don't view that as the hardest part. What I view as hard is, how do you make sure that there's not five ways to do the same thing? Right? Do you need a service for queuing and a service for streaming and a service for notifications and a service—or do you actually need one service or maybe two services that fit together better and you don't have five things?

(Clay at 00:22:49) And so having that kind of foresight of planning out ahead, of knowing where you're going to go, and then thinking through how do you decompose that end state into the fewest number of building blocks you need to make it work together. And then once you have the building blocks, kind of creating a seamless user experience, whether it be through, you know, things like a Terraform provider or whether it be through the actual UI, you know, complex workflows that kind of span individual services. I think that's the really hard part about making a really good cloud platform. And part of it is, it's easy to be myopic. Right?

(Clay at 00:23:24) You know, you hire good people. You have a product manager who works on the compute team, and their customers are telling them, "Hey. I want you to solve this problem." They're typically not going to take a step back and go, "Well, actually, we shouldn't solve that problem in compute. What we should do is we need to build a separate service, maybe it's for task scheduling.

(Clay at 00:23:43) And that should integrate with compute, but it would also integrate with object storage and these other things." And so you have to have this kind of constant curation process where you have people outside of any individual service or product looking at the requirements and what's coming up and then trying to figure out what's the most efficient and effective way to build that throughout the whole platform. That is fundamentally, I think, the biggest challenge we have building a platform.

(Joel Beasley at 00:24:07) So what does your team look like? What is—you have this large organization, but what does your direct team look like?

(Clay at 00:24:14) I think that the first thing probably to cover is what is the kind of organizational structure within those individual teams. So, you know, I run an engineering organization. The way that we handle this is we include product management inside the engineering organization. So you have kind of leaders who have both engineering, product managers, UI people, program managers, all that stuff reports into the same kind of single structure. And then a lot of it is you break it down by—like, my reports, you break it down by functional areas.

(Clay at 00:24:45) So there's a leader who leads, you know, compute and storage and virtual networking. It's a huge space. It all kind of goes together. You have a leader that focuses on data services and AI and ML. You have a leader that focuses on security in the platform space.

(Clay at 00:24:59) Things like, you know, your overall identity and access management, internal security, compliance programs, also security products, things like KMS, HSM, et cetera. And then the same thing across the developer-focused tools. Right? You have a team focused on your UI, your CLI, your SDK, but also telemetry and logging and code repository and build systems and CI/CD. And so you break down the overall organization into those functional areas, and you try to align them based on commonality so that you end up with a cohesive strategy coming out of each one of those pieces.

(Clay at 00:25:34) And then, of course, that just recursively flows down. Right? Within that leader, they'll have a set of people who do X and a set of people that do Y, and then it just recursively continues down until there's actually people who do real work.

(Joel Beasley at 00:25:47) That's the engineer in you.

(Clay at 00:25:50) Yeah. Exactly. Right? That's—

(Joel Beasley at 00:25:51) Devaluing anything that's not code. Yeah.

(Clay at 00:25:55) Well, it's not just code. To be fair, there's design. There's people who actually produce things. And then there's people who, you know, help make sure we're producing the right things. But you gotta make sure you balance the ratio of those people as well.

(Joel Beasley at 00:26:07) Right. That's true. Because, like, they're very necessary. Like, absolutely necessary. But it becomes too much, like, if it's like government.

(Joel Beasley at 00:26:15) Like, you don't want, like, too much of it. Right? You want the exact appropriate amount to make this machine run. No.

(Joel Beasley at 00:26:22) That's cool. Thanks for giving me the topography, like, the overview of that. But you mentioned a couple of these, you know, key leaders. How do you specifically recruit these people? Do you look for them to go to conferences to look from? Do you find them through friends? Like, how do you find these high-level leaders?

(Clay at 00:26:39) Sure. Well, there's a few different ways. One way is you grow them as you go. So as you scale an organization, sometimes you find people that can grow with you, and you promote them and they grow that way. Sometimes, you know the network. Right? You hire people. You have connections, and then you end up hiring people that you know are good and you bring them in. Also, a lot of it is you work with executive recruiters. So, you know, companies like Oracle have a very successful executive recruiting team that have a lot of connections in the industry and, you know, who are people that can handle orgs of these size, and how do you then go and recruit those people?

(Clay at 00:27:16) And then you, you know, you interview them and you sell them on the position and the overall opportunity. So I think you kind of do all of those things. It's not so much "go to a conference." You do look at people who speak at conferences. But typically, you know, you're not going to meet them at a conference because the people we're looking for, they're going to be busy and probably don't want to talk about that right there.

(Clay at 00:27:35) They're giving speeches, and they have a bunch of other responsibilities when you're doing that. But you do look for people that are relevant in the industry and have a perspective and a persona because it's also part of the job. Right? You need, as you become more senior, you need people that can actually represent themselves internally, that can represent themselves externally, that can speak, and that is a part of the job description as well.

(Joel Beasley at 00:27:57) So as we start to wrap up here today, what are you most excited about? What's the big thing? Obviously, this is big news for me with the Oracle Cloud. Like, what's the level deeper? Like, what's the thing you're really excited about today?

(Clay at 00:28:10) Well, I will give you a long-winded answer, a four-part multi-answer. So I think a good question is, what I'm excited about is what we're doing differently here at Oracle. So in large part, you know, I get this question a lot. People ask me, "Well, why should I care about Oracle Cloud Infrastructure? Why isn't, you know, AWS or Azure or GCP? Why are—how are you different, and why aren't those just fine? What—why are you doing this thing?" And so Larry Ellison, our founder, he really talks about this in terms of, like, generation one cloud versus generation two cloud. And so, you know, the engineer in me is forward in my overall market position. So I downplay some of that.

(Clay at 00:28:54) But let me explain to you what he means and what we mean when we talk about Gen 1 cloud versus Gen 2 cloud. If you take a step back and you look at what cloud infrastructure, how it is originally designed for, right? AWS was the first player, and they designed AWS to enable a couple of people in a dorm room to build the next new business. That's really what their overall goal was from the beginning.

(Clay at 00:29:19) And that goal permeates everything they do from the way they constructed S3 and EC2 and SQS all the way up through the most recent services they launch. And then in large part, you know, Microsoft and Google, they tried a different take on the cloud pretty early. I mean, I don't know if you remember Google App Engine or the original Microsoft Azure PaaS layer, but they realized that wasn't going to work. And, you know, I'm not trying to be pejorative, but in many ways, they said, "Okay, we're not going to do that. Let's just copy what AWS is doing. They seem to be doing a good job. Let's go copy their playbook."

(Clay at 00:29:45) When I came to Oracle, what I came here to do and what my team came here to do is to really build a cloud that would enable the enterprise to adopt it. So I'll take a couple of minutes to try to explain what I mean by that.

(Clay at 00:30:05) You know, I think you would agree, in general, everybody wants to move to the cloud. Right? I don't know if you remember a few years ago when everyone still had, like, the dumb phones, you know, the Nokia snake phone. And then suddenly it was like boom, everybody had an iPhone or everybody had an Android. Because what happened was it became very clear that was the future and then everyone just switched.

(Clay at 00:30:26) So I think we're at the point in our industry in the cloud where everyone knows the cloud is the future. But for some reason, we're still only 10 or 15% adopting the cloud. Why is that? Well, in my opinion, it's because it's way too hard to adopt the cloud. The way that cloud providers have built the cloud, the idea is, and you can even hear it in what they tell their customers, you should move all of your stuff to the cloud. And in that process, you should rewrite it to be cloud native. And that's great if you're building new things from scratch. It's an easy thing to do. You write it the right way. Life is simple.

(Clay at 00:31:02) But as you know, the vast majority of the spend, the vast majority of the code out there was not written for the cloud. It was written for on-premise data centers with a certain type of network and certain assumptions about storage and compute and a certain assumption about how to integrate these things together. And customers are really having a hard time moving those things to cloud. And so what we did with Oracle Cloud Infrastructure is we had an explicit goal that we were going to make a cloud that would enable those customers to move their most serious and demanding and hard-to-move workloads to our cloud. Now that's all vision.

(Clay at 00:31:36) What does that mean on a day-to-day basis? Well, there's a bunch of things we did. Very cool things like we have pervasive bare metal technology with very secure white box processes. We have complete off-box virtual networking that other cloud providers have a hybrid mix of those things. We added a whole bunch of very interesting features into things like our block storage where you can do fast clones and coordinated disk snapshots. And we created a lot more features into our virtual networking layer, et cetera, et cetera. The net result though is that every day we wake up with an explicit goal of how do we enable those customers to move to our cloud. But more than that, you know, that's kind of great. That's table stakes. But then we're really pushing on some very interesting new ideas over the past year or so.

(Clay at 00:32:22) So one of the things that is really huge at Oracle is this thing called Autonomous Database. And, you know, Oracle has obviously been the dominant player in the relational database market for a very long time. And when I look around in the cloud, no one's really pushing relational databases to where they could be. And so what's really cool, and this is where, you know, you can actually go sign up with your free trial and try this out, is you get this thing that we call Autonomous Database. And what it means is within a minute or two, you can spin up a full-fledged Oracle database.

(Clay at 00:32:52) You can dynamically scale it up in terms of disk capacity and cores up and down. It automatically patches itself with no downtime. It automatically tunes your queries. We can do automatic index creation. There's a whole bunch of very complex functionality we're putting in the database to really make it easy for anybody to use a relational database.

(Clay at 00:33:12) The thing that we're doing that other people aren't is, I don't know how much you've used things like relational databases. I think they're an amazing piece of technology. The ability to do transactional commits, the fact that you have backup and restore functionality, point-in-time recovery, disaster recovery, replication across different nodes—they're very advanced. However, they sometimes have certain scaling limitations and they can be difficult to use. Well, what other people are doing is they're like, "Hey, why don't we just create 10 new unique databases? Here's a NoSQL store. Here's a graph database. Here's a document store." And then you as a customer, you get to deal with all that complexity. What Oracle said is, "No. We can create a single Autonomous Database. We'll support all those different data types within it. You still get all of the best-in-class, you know, encryption and data recovery and isolation features. And we can offer that to you in the cloud very easily." So that's a huge thing that we're very, very excited about.

(Clay at 00:34:08) And then we're also doing some very interesting stuff around what we call flexible infrastructure. So, as an example, one of the things that we're doing right now is, kind of the state of the art in how you write cloud-native applications is you Dockerize your application and you put it into these little units. And then the idea is you can be horizontally scalable. Right? When load comes in, you spin up more of them. When load goes down, you kill them down. Well, that works great unless your application wasn't written that way. So one of the things that we're doing, and we've already come out with this for our storage layer, but we're coming out with it for our compute layer, is why can't you start with, say, a two-core VM? As load increases, why can't you dynamically add cores on the fly to your running application? And then later, when load goes down, the cores are removed from your VM.

(Clay at 00:34:58) Well, it turns out that technology, if you actually wake up and you do the hard work, you can do it. Our competitors say, "I don't want to do that. You should just carve up your application and spin up new instances." But we're doing that hard work to be able to—so now you can configure how much RAM you want independent of cores. You can scale your cores up and down seamlessly.

(Clay at 00:35:17) We're also doing—we have very flexible block storage where, I don't know if you've used, you know, you end up picking a volume type upfront. Do you want a slow volume? Do you want a fast volume? Do you want a provisioned IOPS volume? And then if you want to change that, it's very difficult. What we have now is you have a single volume type and you have a performance slider. So you can actually say, start off with low performance. It's very cheap. Online, while it's running, you can shift it over, say "I want high performance." The performance is added to the volume, and then you can scale it back down. And so these are significant investments we're making into our infrastructure layer such that you don't have to rewrite your entire application to get the benefits of the cloud. Does that make sense?

(Joel Beasley at 00:35:57) Nice. Well, yeah. It looks like now people are going to be copying you, which makes you the leader. Right?

(Clay at 00:36:03) Yeah. I, you know, I try not to get caught up in who's copying who or who's leading. What I really try to focus on is—

(Joel Beasley at 00:36:10) You got me excited.

(Clay at 00:36:12) I know. I got excited. What are the problems that our customers have and how do we help them solve those problems? The reality is I still remember the very first time I used an EC2 instance. And I realized that I could, like, back it up and restore it. And I no longer had to worry about messing up a computer, and it changed my life. The reality is this is now 10 years later for me, and the vast majority of computing environments still cannot take advantage of that. These are just facts. My job is to make sure that everyone gets those advantages. That's what I come to work every day to do. That's what we're very focused on here at Oracle, and we're super excited to bring that to all of our customers.

(Joel Beasley at 00:36:55) That's like a mic drop moment. This is a great way to wrap it up. Dude, Clay, we did it. We made a podcast.

(Clay at 00:37:00) Thanks for having me on today, Joel. I really appreciate it.

(Joel Beasley at 00:37:03) Yeah. And Jake and Chloe will loop back with PR, let you know when we're going to release the clips and the episode and all of that. And then next time I'm out or maybe the first time I ever go out to Seattle, if I'm out there, I'll send you an email. If you're around, we'll meet up. Say hi.

(Clay at 00:37:15) That'd be great. Come to Seattle. Come in the summer. That's my advice.

(Joel Beasley at 00:37:19) Come in the summer. All right. See you, bud.

(Clay at 00:37:21) Have a great day.

(Joel Beasley at 00:37:21) See you.

(Clay at 00:37:22) See you.