The Lessons Learned From Running Start-Ups with John Puopolo, CTO at CorEvitas

(Intro Narrator at 00:00:03) Hello, my friends. Today we're talking to John, CTO at CorEvitas, and we discuss lessons that are learned only from running startups, what technology might be created after the cloud, and how John's battle scars as a CTO gave him invaluable knowledge. All of this right here, right now on the Modern CTO Podcast.

(John at 00:00:32) Here we go.

(Joel Beasley at 00:00:33) This is the Modern CTO Podcast. So tell me a little bit about you and how you got started.

(John at 00:00:45) I started young. I kind of fell in love with computers when I was twelve years old. My uncle was a mainframe programmer, and at some family holiday, I asked him what he did for a living. And he explained it to me and saw that I was actually really interested in what he had to say. And about a week later or so, he brought me some books from the old John Wiley and Sons series. And one was on an introduction to data processing, and one was about how to program in BASIC. And as odd as that sounds for a 12-year-old to love these books, but I did. And they sparked sort of this lifelong interest in technology. So in high school, I learned Fortran and Pascal and then went on to study computer science in college.

(John at 00:01:27) And then after college, I started my career in financial services at places like—you can appreciate this—at places like Fidelity and Putnam. And after about four or five years of developing the internal software, I moved to Lotus and IBM where I learned how to write commercial, sort of enterprise-scale systems. And while I was at IBM, the group I was working with, they went on to found a startup company, and they recruited me to join them. And after working there for a couple years, I got bitten by the startup bug. And probably for the next twelve, fifteen years, I worked at a variety of startups in roles from senior engineer to architect and across a bunch of different industries, including enterprise search, ecommerce, large-scale digital map making, and location-based services using mobile devices.

(John at 00:02:19) And then about ten years ago, I felt that I had advanced enough in my career that I had something significant to offer a company and wanted to use my time and do something that was mission-driven. And this idea eventually led me to healthcare, and I'm pretty grateful for the opportunity to work in a field that's dedicated to helping to improve clinical outcomes and eventually improve people's lives. So it's been a good journey.

(Joel Beasley at 00:02:44) And what was it about the startup world that attracted you?

(John at 00:02:47) It's a good question. I think it helped me realize what I was made out of. Right? And in the bigger companies, there was always someone you could go to for a lot of help or mentorship. And in a startup, in a few of them, I was the guy. Right? I had to figure it out. I had to make things work. I had to be resourceful. I had to kind of dig a little deeper inside to realize the value in the startup.

(Joel Beasley at 00:03:13) So you're telling me if you push people, they grow? Like you can put yourself in a discipline.

(John at 00:03:17) If you push people, they grow. And more importantly, if you push yourself—

(Joel Beasley at 00:03:21) That's what, yeah.

(John at 00:03:22) And you have to be willing to do that. And, you know, it sounds trite, but it's true. Get out of your comfort zone and push yourself and learn new things and realize that if you work at things long enough and hard enough, you will eventually get there.

(Joel Beasley at 00:03:35) Yes, exactly. You accelerated your journey by going into the startup world.

(John at 00:03:39) Yes, I did. I learned a ton. I probably gained—you know, for every one year at the startup was almost two or three years, I felt, at a larger company because no one hand-holds you at a startup. No. Sink or swim.

(Joel Beasley at 00:03:51) I tell people—because we have less than 15 employees—so I tell people all the time, in the interview process, I set the expectation. We don't babysit. Like, you come on our team, we're all experts. We all run, and you either run and you stay on the team, or you don't run and you're not on the team.

(John at 00:04:05) Right. And I think it helps you understand that you are more resourceful and grittier than you probably think you are.

(Joel Beasley at 00:04:12) Yeah. And then it just helps us get to that next level. Like, I had someone describe the corporate world to me as adult daycare.

(John at 00:04:20) I can understand that. Right? It sounds insulting to some degree, but I understand it. If you only—like, if you go from school to college to, you know, basically a large cube farm where people are still telling you what to do, you never sort of break out and realize what you're capable of.

(Joel Beasley at 00:04:37) Yeah.

(John at 00:04:37) And so I would recommend that everyone do a startup or two at some point in their career, and I think it gives you a different perspective on work.

(Joel Beasley at 00:04:45) Absolutely. And, yeah, there's a huge spectrum. Right? There's the people that go through that process, and then they sort of mentally check out and just show up every day, and they're at the lowest—often, like, the lowest level. And they only get promoted by happenstance. Right? And then you've got the people who are trying really hard. I'd say, you know, at all those enterprises, I don't want to negate the top 10% or the top 1% of people that absolutely crush it, right? But when you have 10,000 people, your top 1% of people, that's a lot of people there, you know?

(John at 00:05:15) That's a lot of people. Yeah, yes, it is.

(Joel Beasley at 00:05:18) Alright. So you went into healthcare data. Correct?

(John at 00:05:20) Yep.

(Joel Beasley at 00:05:21) What are you doing over there? Like, what problem are you solving? How are you bringing value to the marketplace?

(John at 00:05:25) So it's helpful to understand what we do by giving you a little background on what the company does. So the name of the company is called CorEvitas, and it's first and foremost a science company. So what we do is we take messy real-world data about chronic illnesses, mostly in the autoimmune space. And through rigorous methods, including data management, information shaping, advanced statistics, we produce mathematically validated real-world evidence about drug efficacy and safety. So our primary customers are pharmaceutical companies and, by proxy, the FDA. So we report safety issues to the FDA. And we have various business units. So the core is the disease registries, and we have other business units in sort of an orbit around that that contribute to creating datasets. And these include precision medicine. So we have the ability to collect biosamples and then do RNA and DNA sequencing on those and other sort of lab experiments on them. We have a patient experience business where we consult with CROs, or contract research organizations, to develop trial protocols, make them sticky. We have a patient social network called HealthUnlocked, which has over a million and a half subscribers, and it's a community of people with various health conditions that communicate and provide support for one another. And then we have a small division that is a specialty EMR. So over that universe of datasets that we collect, we collect that data, we do some information management, and then we derive insights from that data for efficacy and safety.

(Joel Beasley at 00:07:01) So my brother and stepmom are both physicians, so I grew up around that. Yep. And we always have the conversation almost inevitably every year around Thanksgiving of, when are we gonna have that point where I show up at the doctor and they want my medical records, and I show them my phone and I transfer, I give them access to it. And, like, I hold my history of medical records. I can choose who has access so that we're not doing that thing where we're faxing documents. I'm getting a life insurance policy right now. The archaicness of this life insurance application process, they make a request. It takes thirty days. They print out papers. They mail me a physical paper because they won't accept digital signature. I sign the physical paper, then they say it's gonna be thirty days until they fax it to us. And I'm like, fax still exists? Okay. And it's just been this huge endeavor, and I want it to be really easy. When is it gonna happen?

(John at 00:07:53) I don't think in my lifetime it will happen. I'm in my mid-fifties. I honestly don't think it will happen. I mean, look at all the failed attempts right now. Right? From very large, very sophisticated, and capable companies. You have Apple, Microsoft, Amazon. All of these companies have attempted this, and maybe they're too big. Maybe there's too—you know, a trust factor there. I don't know what all of the hurdles are. I mean, you have political hurdles, you have technical hurdles, you have social hurdles, and it's just such a large problem. Years ago, I looked into trying to approach this problem as a startup with smart cards. Why can't you have a key, a digital key on a smart card that can unlock your records and only you would have that? Right? But then you have to worry about someone stealing that key. It is a very complex set of issues, and they're all interrelated. So I don't know when we'll have that. I get frustrated as well. And having gone into healthcare about ten years ago, it was almost terrifying to see all the different systems and all the different data that flows back and forth, and it seems a little bit chaotic.

(Joel Beasley at 00:09:03) I believe I figured out why, though. I believe that why it's—like, okay. Give me an example. You call to make a doctor's appointment at my doctor, right? And there's, like, an incredibly long hold process. Calling is the only way to schedule. Calling is the only way to cancel. So you're gonna wait on the phone for seven minutes to cancel. And then I had this epiphany while I was sitting on hold, and I was like, they have a captive market. They have so much business. They have so much more business than they can handle. The competition isn't present hardly. And when there's not a financial driver to something, you often won't get the result. Right? Like, if there was financial incentive to have a—like, call center metrics? Yeah. Yeah. Then instantly, we would have it. I say that because the technology is clearly here to have an electronic medical record that you can own and carry around with you. Right?

(John at 00:09:59) Right. Yeah. The storage of that is a problem. Right? If, like, think of DICOM images—and it's very personal data, so you'd have to trust an entity to hold it. Right? Even if it was encrypted, you would have to be the only one with the key, and then the key management becomes an issue. So there are definitely challenges, especially for mass populations.

(Joel Beasley at 00:10:20) Have you on your free time, Sir Tim Berners-Lee, creator of the web, right?

(John at 00:10:24) Yep.

(Joel Beasley at 00:10:24) I did an interview with him a few years ago, and he was telling me about, like, what the internet's going to be in the future and what they're working on and things like that. And these pod concepts where, like, the data resides with me, but I can choose to give you read and choose to give you write access, and I can revoke it at any time so that you can be updating my data store. And I thought that that was actually a really unique way because right now, if you're a customer at Bank of America, the data resides over there.

(John at 00:10:51) Right. You can see where we might apply technology like blockchain to that. Yeah, right? What you own, so it's irrevocable. I mean, you can revoke it when you need to with tokens. It's read-only. You can't mess with it. So I think that could be a technology that we use to solve some of this problem.

(Joel Beasley at 00:11:09) Is anyone doing that?

(John at 00:11:11) Not that I know of. The company I came from, HMS, which was acquired actually by—is now Cotivity and Gainwell—we started talking about potentially using blockchain as an experiment to see what we could do with Medicaid and Medicare data. We never got that off the ground because we were acquired, but it was an interesting concept. So it's in the back of my mind.

(Joel Beasley at 00:11:32) I want to talk a little bit about what do you and your team specifically do?

(John at 00:11:37) We're responsible for all of the technology and platforms that enable our scientists to do their jobs. Right? So data collection, data shaping, data storage, protection, privacy, and then all the software processes that underlie that. So the SDLC, you know, software development, QA, DevOps, IT, data analytics, reporting, and cybersecurity is what my team does overall.

(Joel Beasley at 00:12:08) So you're primarily, like, internally focused to build the common components?

(John at 00:12:11) Internally focused. Right. So most of our customers for technology are internal.

(Joel Beasley at 00:12:16) Okay.

(John at 00:12:16) So those include biostatisticians, epidemiologists, pharmacovigilance folks, and some of those folks are actually MDs as well, doctors.

(Joel Beasley at 00:12:26) What is—I've heard you say it once or twice, info shaping or data shaping. Is that just like enrichment? Is that like a similar thing?

(John at 00:12:33) Right. You take raw data, and then you can either enhance it or reject it based on certain, you know, filters and business rules. You might want to quarantine it. And so you want it—you want it to be slightly—you know, I think of it as raw and then semi-cooked and then curated or cooked.

(Joel Beasley at 00:12:49) Have you come across any systems—I'm thinking that a lot of companies build on themselves, but, like, where you can pump in all your customer data and, like, get a bunch of analytics about, like, who your customers are and, like, how fast deals move and things like that.

(John at 00:13:03) I mean, certainly platforms like Salesforce do that, and then other sort of commercial platforms will do that on the CRM side of the world. For us, we're, you know, collecting a lot of clinical data, and then we want to do bespoke analysis on that clinical data. So getting it to a place where you can do analysis is actually a big effort right now for us because we have lots of different disease registries. We have lots of different kinds of data. And so we have two major projects going on right now. One is the development of a data warehouse to collect the data about all the different disease registries into one place, but that's a highly curated data structure. And so the other major project is building out a data lake. Right? Capturing the data in its raw form, being able to look at it in all its sort of messiness, is very useful in a lot of ways, especially for pharmacovigilance. And then taking that and presenting it in a way that's editable and malleable. And then, eventually, that gets to a highly curated state in the data warehouse. So we're responsible for those two major projects and then all these sort of software development that has to go underneath that.

(Joel Beasley at 00:14:10) How do you keep the focus on the outcome? So, like, when you go down these data projects, at least in my subjective experience, you start out with, okay—I'm just gonna use a customer one because that's what's on top of my mind.

(John at 00:14:21) Sure.

(Joel Beasley at 00:14:21) Okay. I want to know, like, which part of the market of our customers closes the fastest. You know? I want information about who closes the fastest from the deal creation date to deal close date. And you go on this journey, and then you start noticing, oh, look at this. I could do that, and I could do that, and I could do that. And then you, you know, start collecting all this data, and then you have to sort of stop yourself and be like, okay, hold on a second. It's overwhelming. I need to get back to, like, just the original hypothesis, the one thing I wanted to figure out first. How do you keep the focus on that as you have all these massive datasets?

(John at 00:14:55) I think proselytizing. More than anything, keeping your team on task, remind them of the goal over and over and over. And you have to be, you know, vigilant. And if you see it start going off course, you can let it go off course a little bit. Right?

(John at 00:15:10) Because maybe there'll be a great discovery or maybe a great insight, so you don't want to be too restrictive. But I think being vigilant, managing to outcomes, to your point, I think is extremely important and reminding the team what the outcome is. Because it's easy to get lost in the forest with the trees. And maintaining communication, I think, is probably the most important thing, both physical communication like this, getting together, talking about what's going on, reminding the team where we need to be and the value of why we're doing. And I think emphasizing the why is extremely important because, you know, you can talk at somebody for only so long, but if you explain the why, then they develop their own energy around the outcome.

(Joel Beasley at 00:15:55) I want to talk a little bit about clouds because it's a popular buzzword in every part of the market today. But for you, like, the context. What is the cloud doing for you, or how are you perceiving it right now in your world?

(John at 00:16:08) It's an invaluable tool. Right? Because the startup time to either test the technology or investigate a new concept is extremely easy to do and very low risk. Right? So you have a vast array of modern, scalable technologies at your fingertips.

(John at 00:16:27) And before the cloud, most companies didn't have the financial wherewithal to be able to test things out and see if they worked or not. They had no ability to fail fast, really, and so you had to make significant commitments to a technology path that you didn't know for sure was going to work. So I think that's one thing the cloud provides. And then all of these sort of operational overhead, you know, security, setting up equipment, all that stuff is taken care of for you. So the operational burden is removed.

(John at 00:17:03) So you can really focus on your domain problem, and you can focus on, you know, the software, and you can develop the software to solve that problem and not have to worry about the low-level infrastructure. I think that's really important. You don't have to make significant capital investments anymore. You have the ability to scale. You know, as long as you understand how distributed systems work, that's one thing, sort of almost a bit of a downside, really, but not quite.

(John at 00:17:28) It's just an extra burden on the development team to really understand distributed systems because by their nature, cloud systems are distributed. But, yeah, and it's very cost-effective. You know, there are some hidden costs. It's sometimes difficult to estimate what it will cost. Although all the rules are laid out, really relating those rules to real life is very, very difficult.

(John at 00:17:48) Like, what's a compute hour exactly? How would you think about a compute hour? How do you think about how much data is actually flowing? So you have to do lots of measurement to baseline your systems, and then you can start making decent estimates on cost.

(Joel Beasley at 00:18:01) And are you considered a cloud provider in your industry?

(John at 00:18:05) I wouldn't say we're a cloud provider. Not a cloud provider. We're certainly a consumer of cloud. And because we don't offer our data to our customers, so our data is highly valued. It's considered the gold standard in our industry.

(John at 00:18:20) And so we keep all of the data private. And then what we really sell right now is the analysis. Once the data lake is finished, and we're targeted mid next year, we have the option to expand our portfolio of products to offer those datasets for slicing and dicing. And so we anonymize it. Right?

(John at 00:18:42) So we hold no PHI. So we have an anonymous dataset, a longitudinal dataset of chronic illness. And so if we made that available, you can imagine pharmaceutical companies, research labs, research hospitals, all wanting that data and being able to slice and dice it as they see fit. So there is a possibility of expanding the revenue portfolio with data as a service, and then we would probably be a provider.

(Joel Beasley at 00:19:04) That would be pretty cool because I've gotten to talk to some people who've done amazing things. This one guy, I believe he's out of Australia, had taken large numbers of X-rays and used it to predict different things. It would essentially be able to diagnose more accurately than a human by looking at radiographs. That's what they're called.

(John at 00:19:27) Yep. We do some of that now with, we acquired a company called Vestrum, which has imagery of the retina. So we have a lot of retinal images. And there are certain diseases where you can correlate lesions and other—I'm not a doctor, so I'm a little out of my depth here. But lesions on the retina with the progression of certain diseases. And so you can predict and you can analyze and predict disease progression using the retina as a proxy.

(Joel Beasley at 00:19:53) That's pretty cool. I get it now as you're describing it. It's definitely fun to make tools for businesses that ultimately help other businesses. But as you reduce, I'd call it the amount of hops. Right?

(Joel Beasley at 00:20:05) As you reduce the amount of hops to the thing you're going after, the thing that's driving you, you get closer to that problem, then it's more rewarding. Right? So the fact that you're building data that could be directly scanned to help people, help a medical advancement, you're really close to it. And the better tools you build to make it easier for them to do stuff on top of your data, then the better advancements we'll get or the faster the advancements we'll get.

(John at 00:20:31) I think the faster, right? The faster the advancements we'll get and the more in-depth we could get, the more statistics, basically, we could apply to that data per day, per hour, you know, per hour, per day, per week, per month.

(Joel Beasley at 00:20:42) So what comes after the cloud?

(John at 00:20:45) It's an excellent question. It's hard to imagine, right, without just saying that the cloud will grow. You know, if the pendulum swings back like it normally does, there'll be advancements in sort of localized computing, personal computing that'll be extremely powerful. You can see that with some of the, like, AWS edge devices that they have. Right?

(John at 00:21:04) They bring some of the power of the cloud local again. So you can imagine that the pendulum will swing back and forth. That will have extremely powerful local machines again, you know, out in the field doing IoT. I think IoT will be enormous. I don't know where blockchain will go.

(John at 00:21:19) You can imagine that it will have significant application moving forward. You know, if you really project out there, there'll be interesting application once we can have quantum computers and sort of reality of quantum computing. What's interesting, though, is it's not a general solution. Right? Quantum computers won't solve basic problems, but they'll solve very complex problems.

(John at 00:21:44) So there's specialized kinds of problems that quantum won't address. But that's kind of I can see those things going there. Whether it be cloud or not, I think another significant advancement will be in AI, in true AI. Right now, we still have a lot of machine learning. I don't think there's a lot of AI really going on in the true sense of the word or the idea.

(John at 00:22:05) And it's kind of scary what's going on. And if you looked at, like, deep fakes and some of the videos with Tom Cruise and some of those videos, they're incredible.

(Joel Beasley at 00:22:15) I'm not even Joel right now. I'm an intern. We're deep faking this interview.

(John at 00:22:20) You could be. You could be an excellent AI, you know, an excellent deep fake.

(Joel Beasley at 00:22:24) Right? So what do you mean it's not real? Like, it's not meeting the expectation of common AI? Is that what we're getting at?

(John at 00:22:31) I think of AI as beyond machine learning and being able to learn new concepts without having been exposed to the concept originally, only related concepts. So I don't see that becoming a reality for the next twenty years.

(Joel Beasley at 00:22:48) Did you read the transcript of the guy at Google who said that the chatbot had become sentient?

(John at 00:22:55) Yeah. I haven't read that yet because I was so skeptical that I didn't know if I should read it. Did you read it?

(Joel Beasley at 00:23:03) Oh, yeah. It's part of my job. Right? That's the beauty of getting to do this.

(John at 00:23:06) What was your conclusion?

(Joel Beasley at 00:23:08) Oh, I thought it was. For sure. I mean—

(John at 00:23:10) Really?

(Joel Beasley at 00:23:11) Barring, however, if I could—it's like a chocolate chip cookie. It's like you can eat the chocolate chip cookie and be like, yep, that's a chocolate chip cookie. But if you don't know exactly how it's made, if I got to see behind the curtain of exactly how that transcript was put together, then I would have more certainty. But my default was, like, thoroughly blown away. I was just blown away. Forget let's say it's not sentient.

(Joel Beasley at 00:23:35) The conclusions that it was drawing and the stories that it was telling was just, like, it was so cool. Everyone saw the headline, and they almost, you know, do what we do as humans to make life simpler. We condense information to move on. Right? They're like, okay.

(Joel Beasley at 00:23:51) That guy's crazy or that guy's real or whatever, and then you sort of move on in life. And I had classified that, and then I heard from some other people. So I was like, alright, I'll dig a little deeper. And it was, like, I think, like, a twenty to thirty minute read or so, and they had posted the whole thing.

(Joel Beasley at 00:24:05) And I, to this day, tell people it's worth reading.

(John at 00:24:09) Alright. I'm going to read it now.

(Joel Beasley at 00:24:11) I think it's on Medium or—

(John at 00:24:12) You seem like a very thoughtful guy. I'm going to have to go and read it now. Did you see the movie Don't Look Up?

(Joel Beasley at 00:24:17) Oh, yeah. And I thought it was so accurate.

(John at 00:24:19) I thought it was so accurate too. It was scary accurate. Yep. Between that and the—you ever see Black Mirror?

(Joel Beasley at 00:24:25) Oh, yeah.

(John at 00:24:26) Black Mirror. Between those two things. So, yeah, it captured the state of the state pretty well.

(Joel Beasley at 00:24:31) Yes. So if nothing else, we're in, like, the most exciting time. It's highly uncertain, but it's definitely exciting.

(John at 00:24:39) It's definitely interesting and exciting. I will give you that.

(Joel Beasley at 00:24:41) Yes. If we were sitting around in, like, the 1200s in our town with, like, mud huts—

(John at 00:24:48) Well, some of us would be looking at the sky, figuring out the moon and the stars. Right? So—

(Joel Beasley at 00:24:51) That's right. But what's happening around us is just, it's my full-time job, and I can't keep up with the advancements. There are so many massive advancements in every single field. I can't even track them with a team.

(John at 00:25:03) Yeah. I can understand how that's your reality.

(Joel Beasley at 00:25:07) So what is the big takeaway you want everyone to have? Are you recruiting, hiring engineers? Do you want them to know that, like, if they're at a certain size of a company and they're experiencing a problem that they need to come talk to you because you've got the solution? Where do we need to focus?

(John at 00:25:21) I think pharmaceutical companies should pay attention to what we have to offer. We're different from electronic medical records. We have different kinds of data. We augment EMRs very, or replace them depending on what needs to be studied. So the kinds of data that we capture are real world data, which is somewhat different than the structured data that goes into EMRs.

(John at 00:25:45) So that's one thing that any pharmaceutical company that needs a safety study or a comparative study for their drug should pay attention to us. Anyone who's looking sort of to break into health care on the life sciences side, but not quite ready for pharma or sort of a big biopharma company should look at us for employment. I am hiring. Mostly right now, we're looking at data engineers because we deal with so much data and data movement and data processing, filtering, specifically for the precision medicine side, where we sort of sit in the middle of LIMS systems and data collection and labs and have to correlate all that data, analyze it, store it, anyone with experience on, you know, RNA sequence data, DNA sequence data, labs, and how labs are performed and the kind of data you extract from mostly wet samples, you know, serum samples, blood samples, those types of things. I joined the company because, you know, what we did as a company, our mission in health care, the fact that we had a proven and experienced management team.

(John at 00:26:55) So the company was started about twenty years ago, was acquired by one private equity firm, and about a year or so before I got there was acquired by another private equity firm. So we're primarily owned by a company called Audax, and they're a top-tier private equity firm. And so the company is well regarded. It's growing. It's financially stable, and we're in a great position to, you know, expand into other disease categories.

(John at 00:27:22) And it's a lot of fun. I mean, I've been having a blast. And we do a lot of greenfield development. Historically, the company has been so science-driven that technology has taken a real back seat. And that was one of the reasons I was hired a couple years ago is to provide a technology capability to the company, a modern technology capability.

(John at 00:27:40) And that's what I've been allowed and supported to do. And so for me, it's been a blast, and I've recruited a bunch of folks either that I used to work for or from related companies, and I've built a really solid team.

(Joel Beasley at 00:27:52) Nice. As we wrap up here, I just want to make sure that we get, like, a leadership question here. So I like to ask people for their best leadership advice that they've ever received and put into practice.

(John at 00:28:04) Yeah. I probably have a handful of them. That could be a topic for a whole hour probably.

(Joel Beasley at 00:28:09) Well, go ahead.

(John at 00:28:10) I would say, you know, for CTOs and their teams is to take an interest in the business and in the domain that they're working in. The true value of technology comes from solving problems in an effective way. I think it's really important to really dig into the domain that you're in. I think partnering with other executives and thinking across the business versus just the technology landscape, especially at the CTO level. You know, you're a C-level person, and you're just as valuable to the business, if not more so than the CFO, right, or the COO.

(John at 00:28:42) And you have to start thinking in those ways versus strictly about your vertical. I think that's important. And all this comes from my own failures, by the way. This advice isn't, you know, coming from on high. This is because I made mistakes in every one of these areas, and, you know, it's very difficult to be a good CTO and a good executive in general.

(John at 00:29:00) And so I'm always trying to learn and be better at what I do. And I think it's the CTO's job to evangelize and demonstrate the effective use of technology across the business. You can't assume that non-technical people understand the value of engineering and what it brings to the table, so you always have to promote your group. I think also favoring practicality and pragmatism and common sense instead of blindly following the new industry fad. Because if all you did was follow the new industry fad, you'd always be switching.

(John at 00:29:28) You'd be agile, then SAFe, then you'd bring it back to something else, then you'd say, let's try a mixture of waterfall. And the more you do that, the more switching costs that you have. So I think applying common sense to how, you know, to work and to getting things done and to manage the outcomes, especially, is extremely important. I'd say probably lastly, staying current technically is really important and really difficult as you mentioned. You know, building software is part art, part science.

(John at 00:29:56) It's a craft, and it's part engineering, and we're in an ever-evolving landscape and discipline. And so, you know, keeping your intellectual curiosity burning, I think, is important. Constantly reading is important. Experimenting is essential. And being okay with failing as long as you learn from those failures, and then you can apply those learnings to future work, I think, is critical to being successful in this space.

(Joel Beasley at 00:30:23) Dude, that was awesome.

(John at 00:30:25) Thanks.

(Joel Beasley at 00:30:26) You got to write a book now.

(John at 00:30:28) No.

(Joel Beasley at 00:30:28) No?

(John at 00:30:29) I wish I had the time. Maybe someday.

(Joel Beasley at 00:30:31) I could tell after all these interviews, like when people give insights, the tone in their voice changes, the harder the insight was. Like the more difficult and the more pain there was behind it.

(John at 00:30:42) I have my battle scars.

(Joel Beasley at 00:30:43) Yes. This was great. We made a podcast. How do you feel?

(John at 00:30:46) Yeah, thanks so much. It was fun.

(Joel Beasley at 00:30:49) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.