EP. 44 From COBOL to ChatGPT: Modernizing Your Tech Stack with Jeremy Uniake

00:00 - Introduction & Guest Background

Matt Pacheco
Hello, everyone, and welcome to the Cloud Currents podcast where we explore technologies and strategies shaping the future of cloud computing and cybersecurity. I'm your host, Matt Pacheco, Senior Manager of Content Marketing at tierpoint and I help businesses understand cloud trends to better inform their IT strategy decisions. In today's episode, we're diving into the fascinating intersection of traditional business transformation and cutting-edge technology adoption. We're joined by Jeremy Uniacke, Vice President and IT Vice President of IT at Thrive, who brings a unique perspective of evolving from legacy systems to modern cloud native platforms. With over 25 years of experience spanning automotive, telecommunications and digital advertising industries, Jeremy currently oversees a huge technology budget. I'll let him talk to that a little bit. And leads a team across multiple continents.

Matt Pacheco
He's been at the forefront of AI adoption, implementing practical solutions since the early days of ChatGPT, and serves as Thrive's internal AI sponsor. Today we'll explore his insights on practical AI implementation by global privacy and compliance challenges, modernizing legacy systems while maintaining business continuity and all the important things you need to consider when you're modernizing and some of these unique aspects of managing transformation in various industries. So thank you for joining us today, Jeremy.

Jeremy Uniake
Thanks for having me, Matt. I'm looking forward to it. Yeah.

Matt Pacheco
So let's talk a little bit about your career journey. Can you walk us through where you started and where you're at today?

Jeremy Uniake
Where I started, I kind of laugh because I really started a long time ago. I was about 7 years old. My dad got me an Apple II because I wanted a game. I wanted a gaming system. I'd seen computers. I was like, oh, this is going to be really cool. He got me that. And he said, I got you a game. And then he handed me a programming book, which for Apple, for those of you who know about Apple iis, it was written in basic. And he said, if you want more games, you get to go write your own and learn it. And so at that time, I was really curious about how it worked. How did they work? What did they look like? How did they go? How do you make that kind of stuff happen?

I lived kind of in the middle of. I wouldn't say the middle of nowhere, but I lived in a smaller city in northern part of Michigan. And so there wasn't a lot of stuff to do. Right. I grew up in an area where in the wintertime you played hockey, you hunted, you fished, and you were outside and it was cold. So if you did, if you had to be inside, you had to do that so, you know, from that time point, I personally have been interested in computers. How they work, what do they look like, where do they go? You know, when I was younger, I actually wrote a track management software for my track team in high school because they didn't have anything and they didn't have anything on computers. They did everything on paper. And I'm like, this is terrible.

How do we know who did the best thing? You have to go back through, you know, hundreds of things. And once I did that, I realized that's what I was really interested in. I was really interested in thinking, how did they work? How do they go? How do we solve problems? I didn't realize it then. I kind of, looking back, I realized I learned how to solve problems. That's what I was interested in. I was interesting in figuring out what problem did you have that I could solve. So from there I went over to Michigan State University. They were one of the top computer science schools in the United. In the United States at the time I was there. I was there for four years.

I ended up with a dual major in computer science and mathematics and a dual minor in sociology and philosophy. So I really only took four things in college. Philosophy, sociology, computer science and mathematics. That was. It didn't take anything else. Absolutely loved it. It was fantastic. Really got to dive deep in the various areas and have used all of those over time, which is really interesting because my sociology pieces were really related to organizational sociology. So as you move up, you kind of figure out how do you. How does an organization work? How do organizations come together, things like that. But that ties back into your computer science and things. And so, you know, after I graduated from, I went back to Michigan State for my. I started grad school at Michigan State. I always thought I was going to be a professor.

That's what my parents were. So I was like, hey, let's be a become a professor. That looks like a great idea. Made it about a year and went, yeah, don't think. Think I want to do this. I actually want to go and solve problems. So then I went. That's where I started my career. I started my career at Ford Motor Company as a contractor. I was working on their last CAD system. Their CAD CAM system, it was called at the time was called PDGS. It was started in the 1960s. So the system was on version 26.5 and it had 30 different languages as a component of the whole CAD system.

From Fortran 66, which is what the original code was in there was still fortran66 code in there all the way through to high performance Fortran because it was running on some higher parallel systems. And you know, from there I kind of learned a lot of working at Ford and the various things as a contractor. But I get to work on tons of different platforms, different areas. Really dove deep into how do computers work, really down into assembly language and working there, moved from there, became more of a, I went to a smaller company and became more of a consultant. So they started because they had seen the work there and because I had worked in, I don't know, it was 30 languages at the time. It's probably down to about 15.

I was proficient at that time as a consultant having that many different languages and stuff. So then I became a consultant. Worked as a consultant for, oh, I don't know, four, five, six years at a small company all over the place. I worked for the, all of the big automotive companies that was in Detroit area. So you kind of work for the big companies. So at the time was still Chrysler, so Chrysler, Ford and gm kind of different areas because I understood how they worked, what they looked like, how they did. And then other small companies and then right around 2000, Ameritech, which was one of the original Baby bells, needed a consultant to work on. They were building a new platform for. They're yellow pages. So they're print yellow pages. They're building a brand new platform. They needed some expertise.

At the time it was in, I think it was Java and Oracle I think is where it was at. This was a while ago. They hired me on as a consultant. I worked there for a while. I saw the dot com crash coming and went, oh, and Ameritech, you know, as all the big companies do, they do a couple of things. One is they expand, they go, oh, we're going to hire contractors for everything. We're going to go hire all these various contractor companies and they grow and then all of a sudden they go, somebody new at the CIO position comes in, goes, you know what, we don't want to do that. We're going to actually collapse and we're going to bring everything back in house.

And so with the dot com crash and everything else, I kind of went, oh, I can see the automotive companies going through that process and it's usually a five year process. You know, they expand, then they contract, they can expand, they contract. And so I went to work full time at Ameritech and I actually have been part of that kind of history, part of that since then. Over the time, Ameritech got purchased by SBC. It was SBC. Then SBC bought AT&T. Then they bought Pack Bell. Wait, I think it got out of order. They bought Pack Belt and they bought AT&T. So I've been part of mergers and acquisitions. I always thought I was going to leave and go do something else, but I didn't because they always, my bosses at the time always found something entertaining to give me.

So one of the things I worked on was the first iPad application for AT&T. And it was for the yellow page company. So it was for the yellow page company, for our sales reps, because, you know, as a sales rep and for those who don't remember, you had used to have. We'd still have sales reps. We had sales reps all over the US going out selling print yellow pages and then digital yellow pages. And then so, you know, you had a company then. And that's part of what I worked on back then was those things. And then I've kind of just moved through and grown. And every time I would think about doing something else, they would find something more interesting for me to do, including the digital yellow pages.

So went from print yellow pages to creating the versions of digital yellow pages still exist and actually provide millions of leads every year. So it's for yp.com as an example. And that's, that now is our global stack. It's actually, we're actually rolling, we've rolled New Zealand in there, we're rolling in Australia now. And so as I stay with the company, I kept thinking of leaving, kept staying because they kept finding me good new stuff and interesting things. And as I picked up various areas of things to work on, so it would be hr, for example, or finance and kind of building different areas, expanding my horizon, learning new stuff. That's kind of the biggest thing I think is how do you learn? What do you want to learn? Me personally, if I'm not learning, I get bored. So I like to learn.

So give me something new, I'll go play and figure out. And AI is like the ultimate learning environment you could ever wish for. It's not really AI, but we'll have a different conversation about that. And so then, you know, as we go through that, you know, YP was spun off of AT&T. So I learned the how do you spin off a company went through was significant portion of the MA team that was there. And then later on it was YP was purchased by a company called the Other one. So that at in 2017, all of the old YP companies in the United States, the all the ones that were spun off the Baby Bells originally were all combined into one company that was called Dexyp and is now called Thrive. And so from there I've kind of taken ownership of almost most of it.

I don't have all of it. I have the vast majority of it. There's two of us that kind of split activities and behaviors. In 2021, right in the middle of COVID while everybody's at home, we decided to purchase foreign companies. So we bought a company in Australia which is very interesting to try and do an M and A activity and integrate a company that is in a continent that is not doesn't overline with you. Which was. That had taught us a lot. We've done a fair amount of M and A activity since then. That's one of the things that I've been working a fair amount of M and A on. And then we did, we purchased New Zealand after that. So we kind of took over the phone company remnant companies over there.

And then our last activity from a MA perspective is we took on a company called keep, which is a CRM platform. So as we go through and we look at what the history of everything look at, you kind of look at like YP was, you know, if I look through the history of the company, you kind of look at it and you go it's a 110 year old company used to print yellow pages. Like I have a version, I don't have it with me. Otherwise I would show you guys because it's kind of cool. From the 1900, the first print, first phone book was around 1901, 1902, 1900. And so the company itself has been there that long and so but it's changed over time. You know, it really has remained what it is today. Which is a small business champion.

11:19 - Thrive's Evolution From 110-year-old phone book company to modern marketing and sales platform for small businesses

Jeremy Uniake
What is a phone book for? What does it do? It puts the small businesses out there so the small business can get calls. How do you find them? You go look in the yellow pages. Used to be go look in the print yellow pages that then it was go look on the online yellow pages. Nowadays it's kind of how do they get around and that's kind of what we're working on is how to make that available for them. So we have turned into from our perspective, as you kind of change as a company, as you go through and you figure out how does it change same thing with your newspapers and print things. How do you change into this world? We've kind of stayed and become more of a. I guess the term we would use.

I got to make sure I get it right because of the marketing team will yell at me. But it would be, you know, we've become the marketing and sales platform for small businesses. So how do we still do that? We still want to give small businesses lead. So how do you market them? How do you get found online? How do you make sure you have the tools that you need so that you as a small business can grow? Right. Part of that is just getting your name out there. How do you get it out there? Where you put it on things like yp.com, you put it in technically still in a phone book. We still, yes, we still print phone books for a couple more years, not much longer, but we still technically still print phone books.

Neither one of us are target our target users. So then we kind of have, you know, we have the various offerings that are doing that as you kind of work up through the stack. We have what our marketing center, which kind of is that part of how do you get out there? How do you go, how do you, what do you do it? That really encompasses a lot of the stuff that we used to do as a company. You know, give leads, provide leads, but also has new things like search engine optimization, AI optimization. Right, because now you have AI searching, so now you got to do an AI optimization. How do you do social activities? Nobody had to do like nobody had to post on a phone book or anything else.

I mean, maybe you got a note in the paper if you were unlucky or lucky, depending on which way you got there and then you kind of go up from there. So how do you sell? So how do you create pipelines? So now you're taking that functionality that was really levered for large businesses. So you think about Salesforce is really a leverage for large businesses, medium to large businesses. How do you kind of collapse that down into something that a small business can use to make them able to compete? Not exactly compete, but get involved in there, get there and then you have the grow piece. So now you get into the piece of really putting into your CRM doing some automated campaigns, automated marketing things. How do you do all that stuff?

Because you don't have 50 headcount or 100 head count or 200 headcount or 8,000 headcount that you might have that are available to do all that functionality. How do you do that in a small business? That's what we're working on now. That's kind of how we do it. And we take those, we take the ideas and the learnings that we've had from the 110 years. Like, you know, this is how you do it. You take those things because your small businesses haven't changed a ton. You know, your small businesses, like, you guys do this because you enjoy it, right. You don't really probably want to know the technology. I mean, you might want to know the technology a little bit because you're a geek, but. But not because that's what you want to do.

A lot of plumbers, I'm going to use plumbers as an example. A lot of plumbers, they don't want to know how to market to people. They're not going to use their technology. There could be a few, but they're plumbers. They probably went into plumbing because they're interested in solving problems for people that happen to do, to work with their hands and work on things like that, or electricians or any one of those kind of service industries, things like that. They don't necessarily. I mean, there are going to be some. There are going to be some people that love technology, but there's not going to be everybody wanting technology. They're going to want to get the technology out of my way, make it so I can do it and make it as easy as you can for me.

And that's kind of where we're going from this perspective, and that's what we're targeting for. That was a long. That was a long. I kind of went from both history of Jeremy into the history of Thrive, because that's kind of what we have.

Matt Pacheco
Yeah. And it's an amazing journey for both you and Thrive. We'll get into a little bit more about Thrive and the transformation that went through, but also little bit about what you mentioned before. So you mentioned. I was. This really caught my ear. You had a dual background or dual major in mathematics and computer science, and then you had dual minor in sociology and philosophy. Was that the other one? So how has that influenced your approach to technology and leadership? Because that's a pretty unique combination of things, especially with the minors in their sociology, philosophy, like dealing with people the way people think. I'm curious, how does that influence your leadership style or even your approach to tech in general?

Jeremy Uniake
The one thing it always reminds me from approachment, from a tech perspective is that there's always people behind the technology and that there's people using the technology and you can put those people. I look at it from a philosophical. If I look at it through a philosophical lens, it's more of the how are they using it to make their lives better? Right. And that's not true for everybody. It's not true for everything or anything along those lines. But can you use it to make your lives better? If you look at it through a sociological piece and my sociological piece was more related to organizational sociology. So it's really large organizations and how do they form and everything else. It's how is it useful to the organization? Right. Technology is not really useful as technology itself.

There's no anything outside, I mean, except to people who really just want to get down into the nuts and have fun. Right. I mean, you just want to learn. It's good for that. But then it's how do you make it useful to, from a philosophical perspective, how do you make it useful to the individual? From a sociological perspective, how do you make it useful to the organization, whatever that organization happens to be? Maybe it's the organization as your household or is the organization your company, is it your country, your city? Right. You have different organizational levels. And so how does that technology play into those and how do you make it useful? Because I think part of it is you want to do something that's useful, right? I mean, that's part of.

And how does it also doing something for something sake, you know, if it's painful or if it's busy, I mean, like we. I want to work. If I'm going to be doing something for 10 or 11 hours a day, I actually want to make, I want to make sure that I enjoy it, A and B, I want to make sure that it's something that provides value, maybe not necessarily to me, but to other folks as they look at it. And so those have given me different, just different views of how to do it as to how to look through the lens. So like, I think a lot of the things, like at my company, one of the things people are used to me hearing is they'll say, what, what are you trying to do? Give me a use case.

What is the actual thing you're trying to do? What are you trying to, what is the problem that you're trying to solve? Use different words to try and elicit different answers out of them. Because if I say what is the use case to a sales rep, sales guy, I'm going to get something completely different than what I say to a marketer or if I say that to somebody in it, if I see somebody in it, I'll get a very Defined. In most cases, I'll get a very defined, well, you know, very defined, very technical, low level answer. If I ask it to somebody in sales, it'll be like, well, I just want some more leads. Just give me leads, man. Like I click a button and. Or we call it the easy button.

I hit the easy button and the sale was made and I have leads. That's your use case. Right. So you know, then it's. How do you use that term? And so the other parts of those two things. So just still answering your question is in both cases you have different. Like if you take philosophy, you have different philosophical patterns and beliefs. Right. But they still, at the end of the day are trying to answer a question, why are we here? Or what's the reason that we're here? And it's different views of the same thing. So it's different ways of saying something to try and get through to somebody. And a lot of the activity that you should be doing from a technology versus how do I talk to somebody and get. And explain myself in different ways so that I get through to them?

Because they don't have the same vision I do. They don't have the same background I have. They may not have the same, they may be in a different country. So they don't even have the same cultural background. So then that means it's incumbent upon me to figure out ways to explain it in terminology that they understand or that a vision, the thoughts that they have that they can put together, that then they can understand the concept. So it's kind of, it's teaching. Let me get. It's a lot of it's teaching. I mean, I wanted to be a professor. I like teaching. I like explaining, I like people understanding. I like seeing the concept of understanding. So you have to be able to talk to them in different ways at their level.

Matt Pacheco
Yeah, always learning. Teaching too. That's, that's what I do as a marketer. Like I'm constantly.

Jeremy Uniake
You're teaching us how to, how to get somebody's attention, how to change their attention, how to potentially do.

Matt Pacheco
It, how to solve your problems and all that cool stuff. Really interesting. And you guys do a lot of the same thing, sort of for small businesses. How do you get your name out there? How do you enhance the experience? So that's really cool. It all ties together really well. Sounds like you're in a place that's really good for you. Aside from challenge you with your problem solving interest and all that speaking of problem solving, because I'm interested in the, your evolution of your IT systems from, because you've gone from how many. I can't, I couldn't even count how many acquisitions and mergers we've been over.

20:43 - Legacy System Modernization Managing 1,600+ applications

Jeremy Uniake
So. So if I went through it. Yeah, so if I went through it, there's over 10 different companies that have been combined together. So as part of that. Yeah, so as part of that's. We're probably north of. Well, the last count I had was over somewhere over 1600 applications from all the various companies. And it's. I'm sure it's a lot more than that. I'm sure it's a lot more than that. It's just the ones I can actually remember. You know, we're down to under 700 now, which is kind of cool. I mean, that's a lot of what I. A lot of what I. And every one of those things have legacy information. They all look at things in different lights, different ways.

You know, if I ask a perfect example, if I ask the Australian what is an item, they look at me and they go, well, this is what it is. And if I ask a US person, they'll go, oh, this is what it is. And they are not the same. They are not remotely the same. So that means I have systems where people think it's the same, but it's not. And so a lot of what we've done is we have taught through. So how do you know, what did we do with it? And so over the years, we kind of have come up with a couple of us have really kind of thought through what our plans look like. And it's really. We say, okay, take what they have, figure out how to map those things to what we currently have.

Is there anything that we don't have that we should keep or whatever that's still good. And then there will. Then becomes. You start mapping all of the information from those legacy information. You kind of map it to the, what you. What I call the common stack. We call a global stack. Common stack. You try and get to one. I think of the world through the lens of life cycle of a customer. So a life cycle of a customer, you start way over here on the left, which you guys can't see because my arms over here on the left, arm's way over here on the left. And that's where the marketers are involved. So they put something out there that generates at some point in time an mql.

So either a marketing qualified lead or even Prior to that, some kind of activity that they have. And so that then moves through the process of becoming a lead, then it becomes a sale. And then once you've got a sale, then it goes into billing and then revenue recognition. It has all of that activity. So you have that huge life cycle of that customer. And eventually they cancel or they renew or whatever, and you have all these systems that kind of fold into there. So what we've kind of come up with, the way we do this is we go, okay, how do you overlap? Take a company we purchased and I saw, we always start with. We always start with what I call the order to cash systems. So it's the.

How do you get an order in and how do you get cash back into the company? Because those are the things that as a public company, our auditors look at very closely. And by very closely, I mean very closely. So those are the ones that we focus on the most. Because that's kind of the most important ones. If you have different ones of those, you're just duplicating functions everywhere. So we kind of map all of those. We take those systems and the processes, we map them from the old systems and new systems, and then we start with our order to cache and we start going, okay, how do we collapse these down? In a few cases, we go, hey, we're going to take the older companies, the purchase companies. We might take their system and replace the existing one in the current company.

But most likely, because we have set areas and set areas of how we do it, we just kind of map it and then we just start collapsing systems. We also try not to do big bangs. I'm actually in the middle of an almost a big bang. I'm the one that proposed it. I didn't really want to, but we try not to do those. We try to do everything in very nice, smaller sequential steps. Or we figure out ways to do, for example, if you have 50,000 customers in a system, great, let's figure out how to do 10,000 this month, 10,000 next month, 10,000 the following month, 10,000 the following Month. You know, so that you're not like pulling the trigger on day one and crossing your fingers. That's kind of the old school way.

We try not to do that very much if we can get away with it because it uses too much risk to the company. So we basically do all that. We start with our order to cash, then we start with our. Then we go next to our fulfillment systems. Because whatever you're fulfilling, whether it's a print Or a digital online or an ad online or a SaaS product or whatever. The fulfillment systems are usually a lot more complex because they have different things and they're very specific to the products that you have. And then we eventually roll into our, what we call our external facing. So that's kind of the last one except marketing pages, like marketing stuff the marketing teams go do up front. They figure out what they're doing.

I try not to muck with the marketing teams very much except to say your data needs to get into my system so I can play, so we can manipulate it and report on it and everything else. But a lot of that is, you know, it's, it's tough because you have, when you do these kind of things, you have to really get agreement at a very, you have to get agreement at the high level first and then you have to get agreements multiple levels down because everybody's like, nobody wants to give up their system. No matter which way you want to look at it, there's always their system. They don't want to give it up. They want to keep their system forever.

And so you have to then try and get agreement to things that look that way and you really, that has to be done prior to like any type of M and A is a good plan to have in advance to say this is what we're going to do when we go so that you don't get dissuaded by whatever you've happened to find that you missed during the process.

Matt Pacheco
So what would you say are some of the biggest challenges you've faced in all of these companies and all these mergers, I guess modernizing for these legacy systems. Are there major challenges? You sort of mentioned that some like their stuff a little bit. What are some other challenges we face. You face when modernizing? Because I'm sure there's a lot of legacy systems. I mean a company that's 100 plus.

Jeremy Uniake
Years old, 40 million lines of Cobol, like just be clear here, I have 40 million lines of Cobol, some ridiculous amount of lines. The biggest issue is when you do M and A, you kind of have a target, you have a start date and you have a thing. So you can kind of, you like you have a six month to a 12 month window where you can do not anything but it's, you can get everybody to agree. The problem is that, is that after that six month window they have their, you have what I would call your common app and you look at it and in order, excuse me, to upgrade or modernize that system, you don't have to go get all the people who use it on a daily basis to agree.

And the tough part is really getting them to agree because the system does what they think it needs to do, right? It may or may not do more, it may do a little less, but it does what they need. And they've created processes and procedures. So now if you're changing out that system, you actually have to now change out the processes and procedures. That's the bigger problem, is that you have to then get agreement on those. What I, the tact I have taken with the last couple is the stuff that I inherited that they decided was going to be the kind of what I would call the end, like the last of the print systems.

As an example, what we did was, for those cases, what we did is we said, okay, I had a pretty good idea as to roughly where the timeframe was going to be, where it. I wouldn't need them much longer than that so I could look out the future. And I said, look, I think we're going to probably reach an end point here. I won't need these systems after that. So part of that goal then is you look at the system and you say, what do I need to be on in order to keep it? What are the other things I have to do? So as an example, one of our systems was running on HP ux.

And in order to have H, you know, in order to have hpx, I had to have the HPUX boxes and then I had to have this, so I had to have all these custom components, right? So looking ahead, I said, look, if we could get it moved over to a base Linux platform, then I can virtualize the whole thing. I can get rid of all of those HPU platforms. Linux is actually very straightforward. It is supported across the board. I could have a much larger choice of options. And it's, I'm not asking the business to change. All I'm doing is moving is upgrading the base functionality to get us to the point where I can then freeze it outside of the security vulnerabilities and things like that.

So a lot of what I did, or a lot of what we did for the things where I knew there was an end date is I kind of upgraded the technology without messing with the functionality of the end user. So all I had to do would get the agreement was that they would test versus they would test and do everything else. And then people would be like, well, how do you get to the Cloud. And I'm like, well, why would I want to go to the cloud? You have to tell me why I want to go to the cloud. As part of that, but as part of it, because of the concept of how do I want to. How do I modernize it, how do I get to the cloud?

We actually modernize it in a way, we virtualize it so I can dump the virtual machine anywhere I want. So if I want to go to the cloud, I can. If I want to stay in a data center, I can. So I virtualize the functionality that's there. And so that's what. In order to modernize it, I looked for areas without changing the business practices. Now, the downside of that model is that I still have the same business practices. So I'm not getting simplification, I'm not getting any of those things. So if I had complex solutions, complex areas, I didn't solve those problems. So that's the downside of that. Right. The upside is I didn't have to go get an agreement and I didn't have to like blow up the entire world in order to do it. The downside is I didn't get simplification.

So in those cases, usually what I do is pick off very small chunks and just continually iterate on those small chunks. So an example would be, for us, an example would be our commissions platform was 10 years old, and I had. I've replaced multiple commission platforms over the years, and the. Our commissions historically have been very complex. And by very complex, I mean, like, when I try to explain it, people just look at me and they go, what? It doesn't make any sense. So prior to getting to this, I kind of reached a point where I knew we had a couple of problems that were upcoming. We were going to get rid of, out of our headquarters building, which had one of our data centers, and that I couldn't move the commission platform.

So I took advantage of the fact that in the future there was going to be a time where I had to get out of that building and this other commission platform wasn't going to work. So I started the conversation to say, hey, we need to simplify our commission plans. It doesn't have to be easy, it just has to be simplified and we need to move to a very defined way of what our data is and how to do it. And I started that probably two years in advance so that I got them thinking about it as they went through it and figured out how to do it. So then simplification became, rather than simplifying the world. It became one system.

Simplify the inputs to the system so that when we rebuilt it became a much simpler thing so that we could run it in time. Simplification is the big one. It's way harder than anybody thinks it is because somebody can always say, yeah, but there's always the exception to the rule. And it's defining those exceptions to the rule and saying, nope, we're going to handle it the same way every time. That's the complexity you add into coding. That's the complexity you add into AI. That's the complexity you add into everything else. And that's. Simplification is probably the biggest, one of the biggest things you can do to actually make things better. Did I answer your question? I think I hit that one.

Matt Pacheco
Do you have platforms that serve a lot of small businesses, as you mentioned before? How do you balance your innovation, like providing new features for them? We could talk about AI in a little bit. I have a whole list of questions about AI I want to ask you about. But how do you balance like innovation and reliability of the systems for these small businesses? Because I understand, I guess as a small business reliability is probably going to be a big thing. And also keeping up, everyone needs new features, they constantly need new things. How do you balance those things to make sure the small businesses get what they need in the services you provide?

Jeremy Uniake
That is a very good question. And I don't know that I can answer it effectively because. Well, no, and the reason I say that is because it's a push pull, right? Because offering new features means you end up having feature bloat which then becomes something like, how do I then have all of these things and how do I explain it then the usability piece of it, or the uptime or however you want to look at that piece that is better when you have less features. Now they're kind of fighting. In our case, one of the ways we do is we look at it through the same lens as the enterprise piece. So you try to take advantage of the enterprise side, which is mostly where I focus.

You take advantage of some of the enterprise functionality that we've come up with over the years to do things like monitoring and all of that stuff that's there and you bake that into the small product especially. And now that in the cloud based world, it's become a lot easier, right? Because I can deploy it all to AWS and I can monitor all that stuff much easier than I could when you had to Deploy that if you deployed that platform on a laptop that was there. Right. So the AWS platforms or GCP or Azure doesn't either. Any one of the three is fine, make it a lot easier for us to actually figure out what that looks like. But it is definitely a trade off because, you know, the more features you add, the more bloated it becomes. The more bloated it becomes, it's more.

How do I explain it? How do I teach them how to use it? How do I keep it simple? That goes against the simplicity thing that I was talking about before because, you know, as you add more stuff, how do you use it? Like, I mean, there are a hundred buttons in this little screen that's in front of me and you know, I haven't used this tool. I'm sure I can figure it out, but it's, it would take, you know, a little bit of time for me to kind of walk through and say, what does this one do? What does this one do? And, and it's, I think for any software product that is going to be your kind of question is what's the most important piece? And, and what is the, you know, do you live by the 8020 rule?

Or, you know, how do you live? What is your philosophy when you look through that? And that's the product team's job. I will say that's the product team's job. I actually, I have opinions, but I try not to. Try not to.

Matt Pacheco
That's pretty funny. I'm sure one of the things that everyone's asking for is some kind of AI feature because look at every service out there, every company out there is trying to enable some kind of AI feature to keep up because that's what's being asked of them. That's what people are looking for. So this could lead into our kind of AI conversation. How does AI fit into that? What I just asked you about innovation as well, Are you guys looking into that?

35:37 - AI Implementation Strategy Three-tiered approach: core platforms, unique competencies, and prompt engineering

Jeremy Uniake
Oh, totally. So we've been working on AI for the last couple of years. Most of my focus has been inward, not outward because that's kind of where my, my bailiwick has been. I have proposed some things from an outward perspective. We've had multiple conversations about outwards. They're actually taking some of the stuff that I have because I have a different group, I have a different user group. Right. I don't have small business owners. I have people that are at this company working and how do It? So our perspective, my perspective, my new boss's perspective, which is good. A lot of ours are, look, AI is changing so fast. It is changing literally on a month to month basis.

And so some of the ways that you do development or you have done development in the past, like how you do those kind of development things, you kind of have to be comfortable with moving at a little bit different speed, but also looking at a different way. Right. So what we've tried to do is, what I've tried to do is come up with kind of three views of the world, right? Kind of there's the core functionality that you have to have. I just have to have it somewhere. And what is that core functionality? So for us, core functionality is like, we have a lot of employees at the company. I can't afford to give everybody a chat GPT license. Like, it's just not possible. It's not, I mean, it's not cost effective.

So one of our core platforms and when originally, when chat, gpp, chat, tpt, I can never say it, they got to come up with something else. When they came out with their original version, you know, they flat out said, we're going to use your stuff for training. And I'm like, as a company, I cannot, I can't. I own the privacy and security stuff. And I'm like, yeah, we can't do that. So. So one of our core platforms was basically to build our own mini version of it. It's not exactly the same, it's not there, but it costs a hundredth the cost I would be paying, probably even less than that. Think about it. Just using the APIs, we built the screens and we kicked into the right APIs and we're good to go.

It provides a functionality that's already there so that our teams can use it. And we did it really quick up front so we could start encouraging our end users to really start figuring out we need them to play. So one of the things we've been trying to do is encourage people to play. Go find something, go play with it, don't do anything crazy. But like, here's what's safe to play with, here's how you play. Like, I. Because I don't know how to use it the best way, there's nobody to know. Nobody really knows how to use it the best way. They'll say they will. Like all the vendors will tell you, whoa, this is exactly how you do it. Okay, maybe that works for you, but how does that work for the other hundred people over here? Or my Business is different.

So we looked at core billing. So then the second thing we asked was, what are we good at? What do we have? What is the unique thing that we have? Well, the unique thing that we have is we work with small businesses. We track all small business calls for our advertising. So we know, you know, we actually track their calls intentionally to show them how well they've done and everything else. Like, I can tell you, if you advertise here, you got 40 calls and it came from this particular ad or it came from here or whatever else, Right? So some of our core competencies are figuring out where your calls came from. So that means that we have information that we can go look at that tells us information about you as a small business.

So one of the things that we really looked at was, okay, well, we have a lot of calls. We probably have 10 million calls. I probably have 10 million calls worth of data. So then I want to know, what can I tell you as a small business? What can I tell my internal teams the same way? And so one of the things that we really focused on as a core piece, we call it calls of value. Did you receive a small business owner? Did you receive a call value due to one of our things? If you did, what came out of that? And so in the past, we actually used to have people listening to calls and then creating summaries and sending all that information over to the small business owner. Well, we know more about it. So we're north of 90.

I think we're probably up to 98% correct with our AI's prompts. I'm looking at it, listening to it, finding the action items and the values, and then giving that to our small business owners so that they don't forget what they got. So we call them calls of value because that's what it is. You received a call, it's a valuable one from us. We can tell you about it. Here's what it was. You can get it in different ways. You can get it through the platform. You can get an email. It doesn't. We don't care. But we looked at it and said, that's a core function for us. We are really good at. This is what we need to do. So that was the second one.

And then the third thing we really got into was, really, as much as a geek as I am, I want to go build my own LLM. I want to train it, I want to do all that stuff. Yeah, well, somebody else already did all that. They're really good at it. We're not. We're not big enough to do those. I mean, there's probably some of us that could probably have fun doing it, but it wouldn't bring value. What does bring value is really understanding how do you utilize the outlines. You know, looking more at like, prompt engineering. And it's really, truly as prompt engineering, we got deep into that really quickly so that we could figure out how to do it. And then we built what for us, we really just kind of built pipelines utilizing that so that you can go, here's a prompt.

How do you want to use it? Here's a prompt. How do you do it? Goes back to what I was saying earlier about simplification and use cases. What is it that you're trying to solve? What problem are you trying to solve? I need to be not specific, but pretty specific about the problems you're trying to solve. Because if you just say use AI for that, it doesn't work. Use AI to calculate your gl. How can you calculate your general ledger without doing that? But you can say, use AI to validate that the data coming out of this system know system A matches to system B and that those are correct and produce the screenshots to say here, this is correct. So that if your auditors come and look, they go, oh, look, here's where it's at.

I did all this work that can be done by AI. It's a very specific thing, and you can have a human check it. So we got into, here are the base things that we know that we're good at. Then the second piece was, how do we push this out to the business and how do we train? So we really focused a lot more on training. How do you do it, how to use it. We're continually telling them about it. And then we encourage everybody to play. Go play. Here, play with this. If you come up with something, we'll do it. We have a list of, I think we're over 150 ideas that. That we've kind of flushed out and started to say, here's what it looks like.

And a lot of what we're finding is I think, what everybody else is finding, which is it's a fun thing to do with, but you still. It's. It's not a magic bullet. You can't just wave a magic stick and say, AI, I'll solve it. Here's my very specific thing. And then how do I get a human in the loop? Because you still have problems with things being wrong. It's always been the case. It didn't matter whether AI did it or whether a human did it. Some things just are not right. So how do you get a human in the loop? How do you have them test it and what are the manual pieces that you there? It's the same thing we've been doing in it for like the last 20 years.

Figure out what the process is, take a look at the process, where's the automation piece? And then can I have something else simplify that automation piece and then get it in front of a human again. And so it's really, it's smaller building blocks. I mean, we're getting into, you know, we're quote agents and all that, which is really just automate that and then look at it and then figure out how do you do the things that we've done all the time? A lot of the work is really related to figuring out what the heck you've been doing for the last 110 years and going, hey, we need to change that process. Right? That process needs to change. Or we can interject certain areas into that process.

Matt Pacheco
So with integrating AI into a lot of your internal systems for your own teams to use, are there any security or privacy kind of considerations or concerns that you have with doing so? Because that's your bread and butter right there. That's, that's what makes you guys. That data is vital to you guys. How do you ensure that secure and private and doesn't.

43:41 - Security & Privacy in AI Era Data classification, phishing campaigns, and protecting sensitive information

Jeremy Uniake
So, so one of the things that we've done. So one of the things that we try to do is that when I put these things in place, we try and limit the. So when I say play, really encourage people to play with the base systems that we have. So for example, if you're using Salesforce, use Salesforce as AI as an example, right? So it's already embedded in there, so you already, you're picking it off. Or if you want to use ChatGPT, fine. We use Microsoft like most people. Not most people, but a lot of companies we have Microsoft, which means Azure And Azure has OpenAI APIs right into ChatGPT. I mean, it's the same AI as they have. It's just through Azure, which means you get Microsoft Security on top of it.

So I do say play, but then it's also, it also is the, hey, we have certain set things that are there, right? A lot of this goes back to what we have been looking at for years and years, especially when you go into it, how do I prevent data from being downloaded and out. The good news from my perspective, which makes my life a little bit easier, is we technically are still the telephone book company, which means I'm still publishing your name, address and phone number. So the good news is it is technically public. I don't get an easy waiver, but a lot of it becomes. Then what tools do you use in order to look at the security pieces? So it's grab all the logs and then one of the areas that AI is actually good at is looking at security logs.

Grab all your logs, look for it and then start playing. Play with those to say, what are you doing? What are people pulling out? How are they accessing it? And then it's just the same data privacy stuff you had or the data issues you had before, which is look at the data, classify the data, make sure you know where it is and where your important systems are, and then look who has access to it, irrespective of whether AI is touching it or not. And you know, if I put my developer hat on, like give root to everybody or admin I guess, in the world, today's world. But then I put my security hat on them, like, no, you can't do that. But I still want to put my developer give every. Give root to everybody where the heck. There's no big deal.

Admin everybody can have it doesn't matter. I just gave a bunch of IT people heart attacks by saying that when they hear it. You know, security is the big one, knock on wood. So far, you know, what we're seeing are very targeted attacks and very specific ones that are targeting like our CEO and our CFOs and things like that where they're really trying, they're not trying to compromise necessarily all of our systems. They're trying to compromise the really important piece so that they can either like the drift, the sales loft drift problem that happened where people got access to Salesforce, where they then just download all the contacts, right, for the big security companies because that's who they targeted. And so you really. That's what we're seeing right now.

You're either going to get contact information so you can pretend to be them, which then goes back to old school phishing. So it's really into teaching, fishing and you know, that's the same. It all goes back to the kind of the same gut pieces where you got to train your end users not to click on everything, train them to pay attention, don't do that, you know, that kind of stuff. But we're going to see more because we're now starting to see people pretending to be other people with the same voices and everything else. So we've had to institute new things on our help desks and our client care things where it's like you can't necessarily because just because somebody called in, you can't guarantee you know who they are.

Matt Pacheco
You mentioned some of that old school security kind of phishing and all that. What are, what are some ways your team is kind of training and engaging employees on that? Because employees could be a very, I mean even with AI can facilitate them gathering data that could potentially be compromised. What's your philosophy on security training and making it engaging for employees?

Jeremy Uniake
So we have our guaranteed one every year that everybody has to take and they probably skip through, but we also have specialty ones. And so what we do is we try and run a phishing campaign every month. We're using AI to generate the phishing campaigns and then we send it out. And if you fail, you have to take new training and then we, we cycle through the training and then we have follow ups. So you know, and we know who's actually going to get hit by those, right? Like our targeted, you know, our targets are going to be the executive committee, you know, CEO, cfo, those folks. And you know, we kind of, over the years I've worked with them or other people say, look, if you're not sure, don't click. You can always ask me, I'm here all the time.

You can always ask my teams, they're here all the time. And then the other big ones are sales reps because they love clicking on everything. They just love clicking on stuff. But having the secondary class and then, you know, if they fail a second time, then they get a personal call from me as I go. Okay guys, why do you think this is safe? Right? Because it's also true in your real life. Like we've been trying to say, look, this is true for you at home as well. This isn't just me being the idiot IT guy, right? Trying to say, hey, don't do. This is also true at home. You, when you get emails at home and stuff, you personally have to know what a real email looks like because otherwise you're going to get scammed.

And if you kind of put it in that terminology, like this is true for you at home. Your, your parents, your grandparents, your kids, all of those could be our targets. And so we're trying to teach you how to teach them another way of putting it. And that's another way we try coming at it to say, hey, you know, think of your family, think of the closest people to you. Think of your community that you have to worry about.

Matt Pacheco
Yeah, that's very true. It's, it's everywhere. People always asking for Google Play gift cards or can I have your Social Security gift card?

Jeremy Uniake
I have your Social Security number. I just need your Social Security number so I can send you $50 or hey, you know what? Click here and I will give you a $50Amazon gift card to do an interview. Those are the best ones. Those ones collect.

Matt Pacheco
My favorite is type in your credit card number here to check if your credit card number has been exposed.

Jeremy Uniake
I haven't seen that one. Oh, that's a good one.

Matt Pacheco
They gotcha. That's a good one. So let's talk a bit about future. The future where we see all of this going. We talked about some of your past, some of the way you managed kind of digital transformation and all this new AI stuff and security. What technologies, I guess, beyond AI are you most excited for as it relates to whether what you're interested in or as it relates to Thrive and what Thrive's doing?

50:32 – Looking towards the future of AI and IT leadership

Jeremy Uniake
I'm interested in what is the replacement for Google search. Because, Because a lot of us grew up. Grew up. I mean, not grew up, but if you think of the younger generations, like, they could just Google the answer. Like literally you go to Google and you. Or I mean, you could say bang. Or you know, I could probably show my age. Tell you, some of the older Altavista.

Matt Pacheco
Those are either dumb or now TikTok.

Jeremy Uniake
Or TikTok or you know, any of those things. Like, so you would just, you would go to Google. Like, I don't know answer. I go to Google. Right. Which is fine, except now the problem is that with AI, you were flooding the search engines, the, what we thought of as the search engines. We're flooding that with an infinite number of websites that have no value. So the website can't do it. And, And AI has notorious issues with what is facts versus not facts. How do you know? I mean, it's not a. Because it's not AI. It's a probability. It's a probability vector. And. But what does that look like? How do It? We all grew up on, you know, Google kind of made knowledge available to everybody. It truly did. It made knowledge available to everybody. But it wasn't just knowledge.

It was the how do I find the small business down the street, how do I find the local bookkeeper? How do I find where my school is? How do I like all of those things like that just thinks you could just go to Google it or Bing or whatever, it doesn't matter. But when those things are overrun by AI, what's our next choice? Is AI going to become smart enough or an AI version going to become smart enough to do that? Is it not going to hallucinate? How do I, how do I lose that functionality? Do I have to go back and start learning how to use paper maps again? Right. I mean, maybe not quite that far back. But you know what, that's where I'm curious.

I'm really curious to see what that answer looks like, because I think we saw a collapse into kind of the kind of two or three big advertising platforms where you have Google and Facebook, essentially, right? I mean, that's kind of where you have your two platforms. So when those no longer become useful, and I think Facebook's kind of feeling that, starting to feel that pain. I don't know if they're feeling it yet, but like, do you go on Facebook? I mean, it's like ad after ad. I mean, and you don't have any of the people, so why would you go to Facebook? Because, you know, Facebook was originally designed to say, I went there because I could keep up with my friends and family and connections. TikTok's better at that now in some respects to keep up with those things.

So, so what does that look like? What does that next leap? Because if, because if AI gets rid of it, if you have gotten rid of all the use for connections and everything else, then why would I use it then? Then you might be going back to the local bookstore or maybe that comes back king. Really? Bookstore? I mean, I, I like printed books, but I don't know. That's what I'm curious to see. I think, I think us feeding some of the AI models may solve some of those. If we can feed them correctly. I think some of those things are going to be interesting. But that's where I'm really curious to see where it goes. I don't think, I'll be blunt. I don't think we're. There's going to be a Gen AI anytime in the near future.

I don't, you know, there's some interesting things about models. I think the other thing I'm really interested in is when I start looking at multiple uses of models. That's where it starts getting really interesting as to not one giant model but lots of little models where you're taking in your coming very specific. And how well do I get to do those small models that are custom tuned for my use case? Like if I can get really small models that are custom tuned for my very specific use case then all of a sudden there's a world of opening up of doing automations at a very inexpensive way. And so those are, that's kind of from a, that's from a business perspective, from a personal one. You know, I, I don't know. It's exciting. I don't. This is, this is interesting. It just really is.

There's so much to learn. Like I, I spend all day playing and just not playing but reading, mostly reading because I'm on meetings all day but let's be clear, I'm on meetings all day but I spent a lot of time reading, figuring out what does that look like, where does it go, how do we solve it? You know, I also want to know like a lot of the time is spent how does a small business get notified in that case, how do I figure out to make sure that when, that when you open your small business, your little bakery, you're probably not a baker, but you might be, you're a little baker. How do you, how do you then get noticed? You may create a website, but how does that, how do you then, how does anybody ever find you?

How do they know you exist? How does that go? Figuring that out is still really important because you know our growth drivers are going to be those small businesses. It's not the big businesses, it's not the AI companies, it's not those. Your growth drivers are going to be those small businesses because that's what people are interested in. It's really interesting.

Matt Pacheco
How do you see the role of IT leadership changing as AI and automation become more prevalent across all industries?

Jeremy Uniake
That's a tough one. I think it's becoming, I think it's changing because you have people that can do more things, that can be more, that can do more things themselves rather than doing it. You have to be more comfortable. I think you have to become more comfortable about doing little projects and little things like truly becoming. So if you look at it, when we moved into the agile world from a waterfall world, it was really hey do two week sprints and we're going to develop and everything. I think the next phase of that is what I would call project proof of. We call them proof of concepts.

But it's really proof of concept stuff where you can have a single person do a proof of concept and then you're moving into how do I do hundreds of those, if not tens or hundreds of those at the same time? And then figuring out how do I enterprise or scale those proof of concepts. So that means what you're changing is you're not having big, long, massive project plans and everything else. You may have minimal ones, but you have those that you have to do. But you also then still have to figure out, how do I get that knowledge out of the business? So it's becoming wide. You have to become wide is, I think, where you're going. It's even wider than normal.

You have to have a really good understanding of the business and you have to understand that other than that, I think the change isn't. You still have to understand people, they still work for you, right? Are you still there? So you have to have empathy for your employees. You still have to understand what make them tick, what makes them interesting. So you have to be able to provide to that. And then part of the other part of that is really understanding how do you hire for that. Like, you know, because that's the kind of thing if you're looking for somebody who is wide across the board, you're gonna have very. You're gonna have some things that are very specific. You know, if I worked for OpenAI, I'd be looking, okay, there's a few AI model folks that I need, but there's not.

I don't need a bunch of those. Right. What I need is people that are interested because you need people that have, want to learn. You cannot be pigeonholed. I think goes all the way back to the first thing were talking about, which is you have to find people that want to learn and are willing to change on a regular basis. Because everything is moving so fast that if you're not learning or willing to learn or willing to change, then you're gonna, you know, you as a company or a thing are gonna be left behind.

Matt Pacheco
That's excellent advice to both leaders and to people looking into the industry to figure out what's next for them. Question for you about yours career. If you could restart your cloud and AI journey knowing what you know now, what would you do differently?

Jeremy Uniake
Oh, oh, that's a. I think the one thing I would have done differently is that when AWS first came about and everything, I would have found a little bit more time in my day to spend more on the technical side of that piece rather than the older school ones because I think that would have given me a little bit more. It would have given me more insight that I've had to learn outside of there now at an earlier A where I could have kind of looked through it and there's certain things that I could personally have pushed for years ahead of time rather than waiting which means I would have had less work now. Like, I mean from my perspective, I mean like I'm just looking at the work I have now.

Like there's stuff that if I had learned if from my personal journey there's certain things I would have learned probably a few years earlier than when I did then that would have allowed me to then kind of understand where I need giving me a better idea of where I was going to go in the future. So, so kind of understanding the tenants why today? Like I do awful lot of the AI stuff and I started doing that really early because of that. Because of that reason I started down the path really early. So as soon as, when TAD GPT first came out, that's when I think we first started talking about it. It was that early as we started looking through it because up until that point your models were available.

You had ML models that you could do but they were really math intensive and really they were hard to explain. That would be one the other one, I don't know. I, I've enjoyed it. Like, you know, I've kind of, I, I, I kind of go back to the. As long as I'm learning, I'm a happy camper and if I, and then this I would put a second one on there is that I like the people I work with. Like, you know, going to work cannot be painful and that hasn't been a problem for me.

Like I have had, I've had pure luck of working with an awful amount of a large number of really smart folks that I could learn from and then I enjoyed working with and so that means that it's not hard for me to get up in the morning and go sit in my office which is right down the hall. I mean right, literally right down the hall. So.

Matt Pacheco
That'S the dream. Thank you for sharing that with us. That's it for my questions. But I had one more for you because I'm very curious. I heard, I heard you mention it about AI not really being AI. Can you talk about that quickly?

Jeremy Uniake
So when. Yes, when people think about. So, so I went to Michigan State. Michigan State was one of the top five AI colleges back in the, when I went to school there, I won't say when I went to school there, but it was one of the big ones. So there were a lot of AI guys on campus. And so when you talk about artificial intelligence, what does that mean? So I think of me, what do I think of intelligence from a philosophical and sociological point of view? What is intelligent? You know, what does that mean? How do you prove that you're intelligent? You can take it from a mathematical viewpoint. Can you solve a Turing Test as an example? Realistically, when I say, what is AI? It's really the way I describe it, and I've been kind of working on this is. I just.

It's a probability engine. And so what it's figured out is, it's figured out to say, if I look at the way humans interact with the probability that the word, the next word out of your mouth, if you have the previous 20 words, the probability of the next word out of your mouth is this. And I understand that's what it is. So I'm going to put that on there. So then it's. So it's just calculating probability. So if you understand statistics and if you've got into the probability piece, then really what it is, it's just saying that's the most likely thing, the most likely outcome is this. So I'm just going to do it for you. It's a lot more complicated than that.

And I'm sure there's a bunch of AI engineers, they're going to throw things at me all over the place and yell at me and everything else. And yes, I agree, you guys are right. But it's not artificial intelligence. It's that it's learned and looked at it and it doesn't have that additional spark. It doesn't have the new thinking is. The way I would put is a pattern matcher is a probability engine. It matches patterns that it's already seen before and it can consume a lot of data. So it looks really important about that. Right? So you know, it doesn't. That doesn't mean that it can't do things that. That we can do. Right? I mean, I'm not saying that's not, it's just. It's not artificial intelligence. It is not. It cannot think of itself by itself.

Matt Pacheco
Simple enough.

Jeremy Uniake
I don't know. Has anybody else given you a good answer?

Matt Pacheco
Machine learning?

Jeremy Uniake
Yeah. So it is. It's machine learning. It is truly. It's still. Machine learning is. That's the what it is. So has anybody given you answer or if has somebody said we're going to have Gen AI?

Matt Pacheco
I've never actually. No, we haven't really gone that deep into it. I mean, some people have said it's not AI at all, but that it's machine learning. It goes back to the thing. It's really machine learning. So, yeah, that's pretty much it. But that's interesting because that's the. Sounds like the philosophical view of what AI should be. It should be able to think on its own, be creative. I think what a lot of people say about AI is that gives you everything you need. It's usually pretty good, but it doesn't give you that creativity, that kind of human spark originality that you expect from a human. So it kind of aligns like it makes sense.

Jeremy Uniake
It doesn't come up with crazy ideas. That's the way I say it. Like, you know, in order to do like, you know, in order to simplify system, it's too, it is so. But also in order to simplify something, it doesn't have that. What I would call the leap of intuition. Like I remember, it was really, it was one of the ladies I first worked with a long time ago and my wife, between the two of them, what they always said was they're like, the lady happened to be a couple of states over. This was before video calls and everything. She would ask me, hey, can you do this? And you know, my first answer would be, well, no.

And she got to the point where she goes, okay, what I want you to do is think about the problem on your drive home, sleep on it, and then come back and call me in the morning. And my wife actually told me. My wife was heavily into yoga therapy and yoga and she's been doing for a long time. And my wife, you know, kind of said, hey, that's a great idea. You should, you should, you know, add these additional things onto this. You should add this kind of thinking about it and then let your subconscious work on it. And, and kind of over the year and it worked like I would wake up the next morning, I lay out the problem in my head and then I wake up the next morning. Oh, I've got like four different solutions that I call her up.

And so I think of that and I call that the leap of intuition, right? And part of it is AI can do some of that, but there's also that leap of, hey, I've seen this thing. But what if I Skip over these 20 steps and I go, look at this. And I want to be able to see that, and I want to be able to see them kind of look at something that's completely novel, that's outside of what's there. And I think that's the missing piece. It is that spark. It's that intuition. It's the spark, however you want to call that piece that. It'll be interesting to see how we come up with doing that. Right now.

I'm a big science fiction fan, so of course there's another part of me that's like, the other part of me that puts the hat on, goes, oh, yeah, no, that would be a terrible idea. That is not good.

Matt Pacheco
Well, Jeremy, thank you for taking the time to speak with us on Cloud Currents today. It was a really fun conversation, talking about thrive and your journey and a little bit about AI and a little bit about the future. So appreciate you being on with us.

Jeremy Uniake
Thank you. It's been interesting. It is fun.

Matt Pacheco
Well, and thank you to our audience for listening in Cloud Currents. We publish episodes regularly on all the places you can find your. Your favorite podcasts. So check us out there and we'll see you soon.