AI Implementation Strategy Transcript
There Has Never Been a Better Day to Be a Great IT Director
Jay Bradford: How’s it going? This is Jay and Michael, and we are here on the IT Directors Podcast. How’s it going, Michael?
Michael Thomas: Man, I am having an awesome day. How are you, sir?
Jay Bradford: I am fantastic. Look, let me tell you — we have a special guest today. We have Mike Mathews. He is Vice President of Global Innovation and Learning at Oral Roberts University. I mean, I’m excited about having Mike on. Hey, welcome to the show, Mike. How are you doing?
Mike Mathews: Good. Glad to be on. There has never been a better day to be a great IT director or CIO, but there has never been a worse day to be a mediocre one.
Jay Bradford: I feel that. That’s exactly it. Look, and we’re going to talk about some great things with you today around AI and challenges, and I know in your space you have experienced this drastically. And just to your point — hey, AI can help us be better, and it can also magnify our inadequacies in our field. So man, we are so thankful you took the time to join Michael and me today. And just to give our listeners and viewers a little bit of backstory — this is not in our official script, and Michael doesn’t know I’m talking about this — but hey, my man to my left used to work at Oral Roberts University, so he has known you for a long time. How long were you there, Michael?
Michael Thomas: I was there about a decade. So yeah, back through the end of 2011 — about a decade. Mike and I worked together for a little while, and fortunately we still do even after the fact.
Jay Bradford: Man, that is so great. And so just having that partnership with you guys, Mike, and Oral Roberts University — we are so thankful that you came on the show today. Michael has shared so much with me about his time there, and I got to visit Oral Roberts when we were in Tulsa about two weeks ago. I did a tour of the campus. You were not available and we didn’t want to bother you that day, but I came and it was really awesome. So it was just really nice to have you on the show and to get to talk to you about AI and what colleges and universities are doing around it. We appreciate your time today.
Mike Mathews: Oh, glad I could be a part of it.
Michael Thomas: Yeah, definitely.
Balancing AI, Cybersecurity, and Operations All at Once
Jay Bradford: Yeah. You want to kick us off here, Michael?
Michael Thomas: Yeah. I know we talked a little bit previously about AI, and I know that’s something — even from my time when we used to work together at ORU and afterwards — that you excel in. You have been a pioneer of this, which is why I was like, “Hey, this is going to be a great topic to have you joining us from Tulsa, Oklahoma.” So if you wouldn’t mind, can you share with us what it is you wanted to go deeper into today?
Mike Mathews: Sure. So when I think about artificial intelligence, I call it the love affair, right? It’s about two years old — in the way we see it today, anyway. It’s probably 20 years old, but the way we see it today is maybe two years. So you will have all kinds of people on a campus or in your company who are having a love affair with AI, like it’s going to change everything. And that’s great, but the reality is, when you take a look at your operations, you’ve got a 40-year relationship with your ERP system, and you may have a 20-year relationship with your network — because that’s roughly how long Wi-Fi has been around. Every single day I have 50 salespeople bugging me, “Hey, you’ve got to look at the latest Wi-Fi. You’ve got to look at the latest this.” And I appreciate that, but the reality is that you’ve got to balance all three of these things all the time. Because there is nothing more detrimental as an IT director or chief information officer than to get your name in the newspaper because something happened that the world heard about because you weren’t paying attention to operations. And yet over here, people are saying, “Hey, you’re not moving fast enough on AI.” And so I love sharing the story of how finding not vendors that you like or vendors that are affordable, but vendors who are respected in their area of credibility — because there is too much at stake with all three of these balls in the air to risk anything. I want to do a lot of great stuff with AI, but I cannot do it by risking cybersecurity, the network, or our ERP system. So it’s a challenge, but yet it’s doable if you have the right partners.
Michael Thomas: Yeah. We definitely appreciate the relationship that we have grown here, and that is something we are able to provide as well. Even from a partner standpoint, we have been able to see what you guys are doing with AI and even with the Global Learning Center. There’s a lot of stuff. I don’t feel like we have all the time to go in-depth, but can you give us a big picture of what you’re doing now and what that looks like?
Building Six Buildings and Modernizing Your ERP at the Same Time
Mike Mathews: Sure. And so every board of trustees — whether at a school district, a college or campus, or a company — will be paying attention to what’s happening in their industry. They hear about a great operational improvement with AI or some cool things with AI. And so every six months, you as the IT leader have to have something that’s cool, but also something that is operationally more efficient than it was the previous six months. And you have to keep that going all the time. And at the same time, what if you’re in the middle of building six new buildings — which Oral Roberts University is right now? And what if all of a sudden you also have a new cloud modernization project happening at the same time, moving your whole ERP data set into a SaaS model — software as a service model? That is exactly what is happening at Oral Roberts University. We continue to grow, and it feels good to grow — but one bad move can cause a whole lot of problems. And so when we have all of this going on, we are grateful that we have a couple of key vendors who take care of classrooms, take care of the network, and take care of things that there are no real bragging rights over. Because the truth is, as an IT leader, you have to be moving forward and showing that you, as the representative of all IT, are moving forward. Hey, every one of our principals or presidents of a campus or CEOs are watching commercials during soccer games, during football games, and they see all this cool stuff happening with AI, or they see something cool that looks like you can buy anything anywhere just with your smartphone. Not quite true in an enterprise environment. And so great leaders have to take the hit where they can and say, “Yeah, I don’t want to be a naysayer, but at the same time, if I say yes to everything, we’re not going to get where you want to go.” And so I just sent you those priorities by Info-Tech, which are critical because all this stuff is moving at the same time. And how I like to say it is this: have you ever tried building airplanes in the sky? Have you ever tried washing a large bus or vehicle going down the road at 70 miles an hour? That’s what it looks like today being a leader. You’ve got to step back and take a look and say, “Okay, if that car — and I can only get one swipe at 70 miles an hour — how many times will it come around? How do I divide up work that is almost overwhelming to most IT departments today, and not cause them to lose hope, but say, ‘Hey, we’re going somewhere’?” But keep in mind that fundamentally, our IT shop is the best in the country. And so it behooves every good leader to have benchmarks. How do we know that we’re thriving — and surviving, I should say surviving first, and then thriving — amongst other peer institutions that we would consider our benchmarks? And I’m proud to say we are. We just did a study of 11 competitive schools in AI, and we’re number one.
Michael Thomas: Awesome. That’s — that’s incredible.
Mike Mathews: It is incredible, but here’s what I want to challenge everybody with. I didn’t write that myself, because somebody would say, “Hey, Mike wrote it — of course he’s going to say he’s number one.” But the truth is, when AI goes out and looks at everything as evidence, it says it — not me. So that’s pretty cool to be able to say. But it’s also even cooler to be able to say it has nothing to do with finding a silver bullet or having some magical vendor out there. It has everything to do with doing the right thing step by step, precept upon precept. You did your Wi-Fi right, you did your network right, you’re doing your operations right — and you’re not going to get burned that way. But to be at the top is a challenge. Yet it’s fun to be at the top.
Michael Thomas: Yeah, definitely. And I’d say in the middle of building six buildings on top of that, and all of the moving pieces — because there are also people within that organization who have daily meeting requirements and demands — how do you keep your eye on the big picture and move that forward?
Don’t Buy Into the Hype: How to Stay Grounded as an IT Leader
Mike Mathews: Yeah. The big picture is really to always understand your mission and your vision, and make sure you know how to get there by leveraging technology. So I’ve been a fan of leveraging technology and using it transformatively for years, but I’ve never been a fan of people calling it disruptive. I get so frustrated. I’m at the United Nations speaking on behalf of higher education, and I say, “I have never heard a more frustrating term from vendors than disruptive technology.” Who in their right mind would sell something that would purposely disrupt a company, a campus, or an organization? So I think how you stay grounded is really to not buy into the hype, but yet understand where it may land. Be almost a forecaster — ask, “How can we leverage something versus disrupt things?” Because I don’t work in an environment where people enjoy me using the word disruption.
Michael Thomas: Mm-hmm. Yeah. Do you feel like there are common mistakes that you’re seeing in this field? Given that we’re still — I mean, you can make the case that early adopters are the ones using it, and we’re still learning capabilities, and those capabilities are changing daily. But what are some common mistakes?
Mike Mathews: Yeah. I think the common mistake is trying to hear too many voices, right? Because hey, we all hear from marketing, we hear from operations, we’ll hear from students, we’ll hear from all kinds of different camps that have a vested interest in something. And that’s all important, but if you can balance it and say, “This go-around” — back to that concept of washing a big vehicle going 70 miles an hour — “This go-around, I’m going to be able to do this, but I can’t do everything.” And every iteration of something, you’re going to make an improvement. That’s incredibly important. Seldom do I see successful CIOs or IT directors who risked everything, went with the latest grand network, went with the greatest this or that. Rather, I see them saying, “Okay, I get it. I wish I could go buy everything that’s been advertised this week or this month,” but the reality is you can’t. So you had better be listening to the right voices — not all the voices, but the right voices that help you and validate that over time you’re making the right decisions.
Michael Thomas: How do you — from your seat — not buy into the hype? Because everyone is always trying to show each other the latest and greatest.
Mike Mathews: Well, I think the best way is to realize that everybody lives by selling something. I realize I’ve got three network providers trying to sell me something. I’ve got to realize, “Hey, they’re making their living off of it. Don’t take offense, but don’t buy everything in either,” because it may not even be factual. So how can I do my homework with AI to find out if somebody claims, let’s make up the term, a five or six times improvement in Wi-Fi — let me go check on that and find out if it’s even true. Don’t spend a lot of time on it, but I don’t want to be on the bleeding edge of everything. Maybe a few things, but not everything.
Cybersecurity, Data Privacy, and the Guardrails Every Organization Needs
Jay Bradford: No, you brought up a good point, Mike. You talked about security and being on the bleeding edge and taking a methodical approach. But security is one of those things — especially around cybersecurity — where hey, we don’t want to be on the bleeding edge as a CIO. I was in your seat before in the local district here, and with security, you want to be very methodical, very thoughtful, and very slow in making decisions that are going to impact students in teaching and learning. Because at the end of the day, you’re responsible for all the students, all their data, and all your faculty’s data at Oral Roberts University. That’s not a small task. And then you factor in AI, where people are like, “Hey Mike, we have this cool new AI dashboard. Can we connect AI to our ERP system? Can we connect AI to our student information system? Can we connect AI to where we get test scores?” So those are things I’m sure you’re asked all the time, and you have to kind of say, “Hey, let’s pump the brakes. Let’s vet this. Let’s put some guardrails around this.”
Mike Mathews: Oh, absolutely. So we’re fortunate — about four years ago, I led the team that came up with an AI policy. Because back then, four years ago, people were talking about AI ethics and just a bunch of flimflam — excuse my language — about ethics. And I said, “Hey, hold on a second. Every student whose parent will lend them the car, or who is maybe able to buy a new car, has AI within that vehicle. Are we going to stop that vehicle from coming on campus because it has AI in it? No. And so let’s not get carried away with AI ethics. Rather, let’s have a policy statement of what we believe AI can and should be used for.” And the first thing would be our president saying, “Let’s be real here” — students who graduate from Oral Roberts University will be expected to use AI no matter which field they are in. So if we are hindering the ability to use AI, we are hindering their ability to be great employees who go out and impact the world. And so ORU is privileged to be able to say we have a 99% placement rate with all our graduates. They either go to graduate school or they go into a career in their field of study — very successful careers at that. But it comes from getting the basics right, not trying to impress people with technology, but using it for what it was intended to be used for. And don’t let that fool you — we’ve got some cool stuff happening too, but we are able to say, “Hey, this is some cool exploratory stuff — it’s not fully operational yet,” and separate those two. Because I also lead the innovative area for ORU.
Michael Thomas: Yeah. I think it was over 10 years ago — when did the GLC first open?
Mike Mathews: The Global Learning Center opened 10 years ago.
Michael Thomas: Okay. Yeah. I remember going in there and they had a 3D lab, and I didn’t see anybody else doing this. You could get in there and take an engine apart or work through a lot of different use cases. So I know we’re talking to somebody who has been utilizing this and has been on the front end for a while. But I also appreciate your insight and perspective into how you weigh using this wisely and being intentional about how you’re doing it.
What AI Operations Actually Means — and How ORU Cut an 11-Day Process to 2.7 Seconds
Jay Bradford: I mean, you talked about just being an innovator in this space. Hey, I’m a heavy adapter and user of AI in my role here at Clear Winds supporting our sales team, engineering team, and marketing team. Just this morning we had a demonstration from our marketing team about how to use an AI tool to make some of our salespeople’s lives easier. And so these tools are so great. To your point, a college graduate and a student has to know how to use these tools if they want to be marketable in today’s current climate. So why should we hold them back? We have to empower them, and that’s what’s so great about what you’re doing at Oral Roberts. It really excites me to know we have universities building this framework for our young students. But can you break down what AI operations means and what makes it different from chatbots, AI assistants, or artifacts? When someone says “AI operations” — what does that mean?
Mike Mathews: So here’s what I believe it means: every entity that you’re managing IT in is called an enterprise. It’s got a starting point, it’s got an ending point, and there is not a piece that is not tied to some form of data, a data table, or some analytics. And so if you’re going to use AI operationally, it can’t just be, “Hey, I improved something over in one area.” Three years from now, it may affect other areas in ways you didn’t see coming — but it may only take three months. And so taking an enterprise-wide approach and saying, “How can we set a timeline for the entire enterprise and say, ‘Here’s what we ultimately need to improve'”? I think if people step back and take a look and say, “What is the number one operational thing that requires the most rework, requires the most time, and prohibits us from pleasing our customers?” — those are three basics. And so we did that two years ago and said the number one thing is transcripts coming from high schools and other entities. How do we take that transcript, remove all the manual touchpoints that are approving it, and speed it up? So we went from an 11-day average transcript validation process — which included a lot of human touchpoints — and turned it into 2.7 seconds.
Jay Bradford: Which is incredible.
Mike Mathews: From start to finish. And so that was good. But now keep in mind — there was probably two years of saying, “This is our number one issue. This is our number one issue.” And so when you start understanding the issue and you can document it through flow and say, “Yes, indeed, we can change things and improve things” — that became our first success. As the IT leader, if that had been a failure, I’m going to have a very hard time convincing anybody to do another AI operational improvement.
Jay Bradford: That’s a great example.
Michael Thomas: I think the key thing as well — and it’s a conversation we’ve had many times — is understanding your own workflow and your own processes. I think a lot of times when you’re working with different departments or entities, they say “there’s this big issue,” but hey, what actually feeds that, and how did we arrive at this point? If you understand the different variables, now you’re in a position where you can actually answer the problem. So that’s —
Jay Bradford: No, that’s huge. You know, Mike, you talked about exactly what I’m doing internally in our role around AI. We have a director of development for applications, and James and I are working together almost daily, Michael, on our AI framework for organizations — to help them get to a point where Oral Roberts is right now. And it comes with investigation. You’ve got to be thorough. Like you said, “Hey, what is our main issue? It’s taking us 11 days to process transcripts.” Now you’ve got it down to 2.7 seconds, so you’re saving thousands of hours. Now your clerical staff at Oral Roberts can do more innovative things — like, “Hey, what kinds of classes do these students need? What kind of track do they have?”
Mike Mathews: Now, what’s important to note is this: if I had gone to the executive cabinet or other parts of the campus and said, “Hey, we’re going to put a chatbot in to help you” — that’s a negative.
Jay Bradford: Mm-hmm.
Mike Mathews: Instead, I said, “We’re going to improve the entire process.” Now, little did people know we were going to put six bots behind that.
Jay Bradford: Yes.
Mike Mathews: The key really is — don’t get lost in the language that puts up fences. As soon as you say “chatbot,” people already have a paradigm about that. Is that an AI agent? Is that agentic AI? Is that this or that? Just say, “We’re going to improve the process,” and wait until it’s done. When it gets done, you have to be your own recorder of this. What I did right away — because I know there will always be someone to say, “I don’t think it improved it” — so I did a video of the old process, a video of the new process, and said, “Here’s the time breakdown.”
Michael Thomas: That’s great.
Mike Mathews: Yeah. And so you’ve got to own it like you’ve never owned anything before, because there will always be naysayers — whether I work at Google, at a university, or at a high school. That’s just the reality.
Michael Thomas: Because truly — you want to just empower them to do whatever it is that’s their main focus, their main goal, right? Our job is to empower them and set them up for success. If you get into the weeds sometimes, it’s hard to get that buy-in or introduce that type of change. But when you can speak the language of the pain point and present that solution, you can’t deny it.
Access, Trust, and Choosing the Right Vendors for AI Implementation
Jay Bradford: No, you can’t. And Mike, you kind of explained the whole process of how we should implement it. You walked us through it from start to finish with that thoroughness and investigation. So what kind of access and capability do these tools need in your environment to achieve AI operations?
Mike Mathews: Excellent point. So if it is truly going to be enterprise-wide, your in-house team — if they built it — or your vendor is going to need almost full access. Otherwise it’s not truly across the enterprise. So we’re back to the real challenge: can you find a vendor who you have proven to respect — not just like — and trust with key things? And we’re very picky about that. I mean, truth be told, in a perfect world, we could find one vendor who did everything for us. But the truth is they all specialize. Some specialize in network, some in ERP, some in AI. But truth be told, it’s just risky. The more a vendor can own within your environment — one who you trust and respect — the better off you’re going to be.
Jay Bradford: No, that’s a great point. Because you have access to very, very critical data — Social Security numbers, addresses, demographic data, test scores, pay scales, bank routing numbers, account numbers. So there’s a lot of very critical data that, when you want to do AI operations, you have to be careful. Hey, I had a meeting with our VP of operations and we were walking through some dashboards of how we could help internally, and I said, “Hey, I’m going to need that spreadsheet you keep referring to.” Well, on that spreadsheet is people’s pay, their deductions — and I’m like, “Well, to work with the data, we have to have that, right?” And so to your point, I’m internal, so that’s fine. But from a vendor coming in saying, “Hey, I’m going to need all your financials at Oral Roberts” — you’re like, “Hold on a second. You haven’t even logged into my network yet.”
Mike Mathews: Yep. Exactly.
AI Architecture in Higher Education: Enrollment, Staff Shortages, and the Future CIO
Michael Thomas: So big picture — how would you say AI architecture fits into the broader role in higher education? I know we’ve got the different aspects of enrollment declines, staff shortages, ticket backlogs. How does it fit into that broader landscape?
Mike Mathews: Sure. I mean, there are many ways it’s going to fit, but let me start at the highest level possible. I think any time there is a technology like the smartphone, like the internet, now like AI — it gives the opportunity to reimagine the future, right? So anybody out there should be reimagining: what if this pans out and it’s going to be as good as we think it will be? What does our future look like? How should we be changing? That’s the first step. Because the truth be told now, the numbers are coming out that all this money people thought they were going to save with AI replacing people is not a reality.
Jay Bradford: That’s exactly right.
Mike Mathews: Yeah. And the price keeps going up in these AI platforms, right? So let’s just take a campus our size. For us to get Microsoft, Gemini, Anthropic, Groq — pick your platform — it’s about a $300,000 request to give everybody access. Not every place is going to say, “Okay, $300,000 — what does that really equate to? And what if it’s the wrong platform? Who’s going to be in business two years from now?” So you really need to have a thinker out there who can say, “Okay, I can think both sides of the coin. I want to see the best of everything with AI, but I’ve got to be realistic and say I don’t want to bring the organization down, nor do I want to take a risk we don’t need to take.” And so you’ve got to start at that level. Moving down the list, you can start helping employees. Because I can’t go to a place of employment today where, if I said to everybody there, “Who feels overwhelmed at work today?” — every hand goes up, right? The smartphone has never really bought us more efficiency, unfortunately. It has caused more work. And that’s why it’s like, hey, everybody who sold it as a disruptive technology was not being truthful. Every network company that has promised server consolidation has not delivered. None of this stuff has materialized. Now, how do I know that? Well, I’ve got this position where I’ve got to be watching the percentage spent on IT. For 25 years, IT spend has never gone down. All the server consolidation promises, all these promises — none of them have materialized. But the same theory would be true of your smartphone. Better deals through T-Mobile, maybe AT&T — whatever you want to look at — at the end of the day, you’re still spending the same. These companies have to make money. And so in order to start rationalizing this and say, “Okay, what are we really trying to achieve?” — one, efficiency gain; two, more customers, better satisfaction. Reduction in workforce — that’s last. But people will always fear the worst first. “Oh, they’re trying to replace teachers. They’re trying to do this or that.” Years ago — let’s say about five years ago — I was invited to go to Bangalore, India, because a major scientist over there had discovered they could use robots to replace teachers. They came to Oral Roberts because we were using robots even back then. And he said, “I’m going to replace all the teachers.” And I said — General Ray was his name, he was the principal of this place — I said, “General Ray, listen. You may think that theoretically you can do that, but it will never take place, and here’s why.” And I told him why. So he invited me to come over and help them understand this. Here we are eight years later, and nothing happened.
Jay Bradford: Mm-hmm.
Mike Mathews: Except a lot of research. Because the truth is that any of these AI tools or network tools are not replacing a person at all. The goal is to replace and improve a system. That’s the ultimate goal. So when you can depersonalize it, it helps everybody. Taking it personally is the wrong approach. My job is not to do a cloud modernization project to reduce people. It’s to really make our customers happier and keep us in business long term. I like using the example of Amazon. Amazon has been using AI for years, and they have proven that the more you improve, the more employees you hire. Because why? Everyone is a Prime customer. How did they get people to want to be a Prime customer? They leveraged logistics and the art of managing logistics very well. They pleased the customer. Now, can every university, every network company say, “I want to be a Prime customer of yours. I love you so much, I respect you so much, I have so much trust in you”? That’s where we really want to get with leveraging the right tools. But here’s the challenge — on our campus, which is true, we’ve got 67 different systems that I’ve been managing for a couple of years now. It’s okay, but it’s still a headache. Now all of a sudden, imagine I’ve got 67 systems and 5,000 AI agents. I just changed my world. I almost cut my own throat by giving the promise of all these little nickel-and-dime AI things that look like they’re going to improve something, but now they’re in my environment and I’ve got to manage all of them.
Michael Thomas: Mm-hmm.
Mike Mathews: I believe in the future — and those of you listening may not know this — the first recorded case of somebody hiring a chief information officer was in the year 1980, by Boston Bank. Within 10 years, everybody needed a CIO. Whether they did or not, I’m not sure. I believe the future of the CIO — it has been around since 1980 — my title is not even that, even though I now own all those functions. I’m the Global Vice President for Learning and Innovation, which includes that function. So the CIO role is just a function. I think we are going to hit a day where we need AI maestros — somebody who understands systems well enough that it’s like conducting an orchestra of 5,000 AI agents and 57 systems. But which systems can ultimately be replaced, overridden, or have a wrapper put around them — pick your method. But it’s going to take somebody who is far more knowledgeable than, “Hey, I got my Microsoft certification, so therefore I’m a genius in technology.” And unfortunately, every CEO has a cousin or a friend who took some kind of servers or SQL training, and they all think they’re an expert now. That will shake itself out in the future. So I mean, the need for top leaders has never been greater in IT.
The Rise of the AI Maestro: What IT Roles Will Look Like in 5 to 10 Years
Jay Bradford: No, I think you are spot on, Mike. When I was listening to him talk, Michael, it came to mind — remember back in the day? I’ve been in IT for over 25 years myself. Started out as a network admin, network tech, PC tech, programmer, developer. I’ve done the whole spectrum — virtualization, data center. And when I started out, Mike, there was a role — you probably won’t believe this — where all a person did was create network logins. That was the entire job. They were called the network administrator, and they got paid very well. This was like early ’90s, and I had buddies who were network admins. I was in college graduating with a networking degree, and that’s all they did all day — create user accounts and delete user accounts. That’s all they did. And now look — that’s one small function to Mike’s point of what a help desk tier one person handles — setting up your laptop, you know. These roles have changed, and it’s really incredible listening to you talk about how that has morphed. And I agree with your methodology about orchestrating these AI agents. Like, I think about myself — I’ve created four or five internal tools, man, and I’m thinking, “Man, we’re saving time, we’re doing great things.” But now it’s just creating more work because we’re going to the next project, because the speed of business has increased. We’re able to do things much faster and deliver information much better to our customers. But hey, now what are we doing next? So you’ve got to have someone orchestrating those agents, like Mike talked about. I think for our listeners and viewers, I really want people to hone in on what he just talked about. Because if you’re at Oral Roberts right now and you’re a student, I would learn every LLM. I would learn how to manage these agents. I would learn how to do the data dashboards. Because that’s going to be the role in five and ten years — managing all these agents. That’s the new help desk.
Mike Mathews: Oh yeah, absolutely. You know, when you look at this and you say, “Wait a minute — how do I really step back? And could I, in my possibilities, see a better workforce that’s not so stressed out?”
Michael Thomas: Mm-hmm.
Mike Mathews: Could we orchestrate AI to that level? Because the truth is, everybody is worn out — fatigue is everywhere. And so that’s my hope for AI one day: that it is actually able to solve some of these big problems that we’re struggling with, without just going after the nickel-and-dime of another agent, another this, “I spent my budget, it’s gone.” Instead of stepping back and saying, “What are we ultimately trying to accomplish?” And it’s frightful, because truth be told, these are trillion-dollar companies now defining the future.
Jay Bradford: Yeah.
Mike Mathews: This is no longer IBM, Microsoft against Apple. Those days are gone. These are trillion-dollar companies passing money amongst each other to build big data centers that may not even be needed. And there’s a company down in Austin, Texas now called Black Swan — they have proven that if it wasn’t for Oracle, the data centers wouldn’t even be needed. The throughput required to manage an Oracle database and some of the other big databases is what requires those big data centers. And so it’s like, wow — it’s going to be interesting to see how this all plays out when trillion-dollar companies are sort of winging it a little bit.
Jay Bradford: I read an article just the other day, Michael — I hadn’t shared this with you — and Mike, in the state of Utah they are building a huge AI data center for large computing and processing. It is going to generate more power than the entire state of Utah does right now, and they are trying to figure out how in the world they are going to support it. It’s going to require what the whole state currently uses — and they already have power issues and water issues in Utah. And so to Mike’s point, there is some real concern around that. There are some guardrails that have to be put in place. But these things are never going away. Like you said, these trillion-dollar companies are only growing. And so it’s very interesting to hear your approach because you’re at a high level — you’re seeing all of this.
Michael Thomas: Mm-hmm. And that informs how you’re taking these steps as well. I don’t hear that with everyone, but I want to call that out — we’re going to champion our mission and our vision as an organization and find ways that technology can efficiently help us move forward.
The Science of Technology vs. the Art of Using It
Mike Mathews: Yeah. I’ve often told people — maybe five years ago and even more so today — the science of technology has been pretty much proven. Moore’s Law, how it increases and so forth, how you can hire almost any company to service things. But what has not been proven is the art of using it. So take the science and the art and start realizing, “Okay, how do we artfully use what is available?” Not always free, of course — but at the same time, let’s test at the right time and then roll it over at the right time. So it’s really: how do I be innovative and realistic at the same time? It’s not easy, but you still need both.
Jay Bradford: No, Mike, I think you’re spot on. And I remember five and ten years ago, the big thing in K-12 and higher education was one-to-one, right? Putting a device in every student’s hand. But now you don’t really hear one-to-one as much. They’re going to carts and labs and different things, because we’re learning that you can’t just stick a piece of technology in a student’s hand and expect magical things to happen. So um, just in closing, Mike — if a listener wants to take one practical step, like “Hey, Mike at Oral Roberts just blew me away and I’m fired up. I want to go home and do this in my organization” — if they want to take one practical step, what should they do to implement some AI?
One Practical Step for IT Directors Ready to Act
Mike Mathews: You know, try and figure out what your board is looking for. Have an ear to hear and be in alignment with them. But it won’t be far off from your mission and vision. So if you can find what is that one thing that allows your board to believe you’re doing something meaningful — which is very easy to do with AI — but then what operational thing did you improve that really makes people believe you know what you’re doing, you’ve got a handle on all your systems, and you can do end-to-end AI implementation? That would be my number one recommendation. Open your ears and be in alignment with your boss, your boss’s boss, and then the shareholders — which more than likely is a board of some kind, a board of trustees or governing board, whatever the case may be. And don’t listen to all the noise down here. Now remember — you do that, and then all these vendors talking to you, it’s like, “Okay, yeah, I get it.” But the fact is, what helps me with my board is when I can find a vendor like Clear Winds and say, “Hey, they get what our board wants.” And it’s not just because Michael worked at ORU for 10 years — it’s because they get it, they’ve listened, and they’re in alignment as well.
Michael Thomas: I appreciate that. I think that is one of the things that we do. Even from this conversation, we want to be in alignment, we want to understand what is that overall purpose, that overall need. Your whole goal is to help that move forward. Your whole goal is to meet that pain point and address what it is. It’s part of our mission to help empower others so that they can focus on and fulfill their own mission. So I think that’s one of the reasons why we play well together and work very well together.
Jay Bradford: That is fantastic. And Michael, in closing — you talked about one thing numerous times, which is not implementing the wrong tool and creating more issues. And I think that’s where we do well at Clear Winds — really partnering with our customers. We want to understand what they need, how they’re growing, and what they’re looking like in five and ten years. But the worst thing we could do is implement something that’s going to make your life more difficult and then make your stakeholders’ lives more difficult and then no one is happy. And I’ve seen that happen with AI. And that’s where I think your approach is fantastic — I’ve loved this episode, to be honest with you. I’ve done a lot of podcasts in my life, and this has been a great one. I want to go back and listen to it myself, because some of the things Mike talked about I think are really groundbreaking in terms of methodology and approach. Man, we cannot thank you enough for being on our show today. I think our listeners and viewers are going to love hearing about AI operations, agentic AI, and what you’re doing at your university globally — not just locally. So hey, on behalf of Clear Winds and the IT Directors Podcast, we are so thankful you came on. And to all our listeners and viewers, go check this episode out. Go follow us. Go check out Mike Mathews at Oral Roberts University. Go follow him — I’m sure he puts out great content on social media. We appreciate you so much, and as always, thanks for joining the show.
Mike Mathews: Yep, my pleasure.
How AI Can Help Your Business: Start With Strategy, Not Silver Bullets
If this conversation with Mike Mathews proves anything, it’s that how AI can help your business has less to do with finding the latest tool and everything to do with building the right foundation. From cutting an 11-day transcript process down to 2.7 seconds to managing 67 enterprise systems while building six new buildings at the same time, Oral Roberts University’s approach to AI is a masterclass in what it looks like to lead with mission, move with intention, and choose vendors you truly trust. The IT directors and business leaders who will win in the next five to ten years are not the ones chasing every new agent or platform — they are the ones who understand their board’s priorities, document their biggest pain points, prove their wins with video evidence, and position themselves as the AI maestros their organizations need. For more conversations on how AI can help your business, subscribe to the IT Directors Podcast on Spotify, LinkedIn, and Instagram, and visit clearwinds.net for show notes and resources.
