20-Person Company That Operates Like 200

Auto-generated from the episode audio and lightly cleaned up. Expect the odd misheard word.

1. 0:00
   Wes Mathews

   Stephen Wojno. At 26, you became a shareholder of a bankrupt company and took it from 7 million to 28 million. Every visionary, every CEO, anybody and everybody that I talk to, everything is about AI.

2. 0:14
   Stephen Wojno

   If you manage great people or lead great teams, you know how it feels to put someone in charge who will get it done, get it done right, and get it done without drama.

3. 0:24
   Wes Mathews

   This is bigger than the internet. Right.

4. 0:26
   Stephen Wojno

   We can now bring the technology of a 200 person company into a 20 person company.

5. 0:32
   Wes Mathews

   But I see a lot of visionaries stuck with AI and trying to figure these things out. You just need to know it's power and how to harness that.

6. 0:42
   Stephen Wojno

   You must start in 2026 because we're not turning back.

7. 0:47
   Wes Mathews

   It's only getting smarter and faster. This is the unfiltered truth about entrepreneurship. Raw, no BS, no sugarcoating. Welcome to Entrepreneur Intel. I'm your host, Wes Mathews. Each episode we'll learn from experienced founders and uncover the top 5% learnings that led to their success in all things personal, family, and business. This show is sponsored by Stealth Consulting, delivering clear marketing strategies, ROI, and no surprises. I am excited for today's podcast. So it's the first podcast of 2026, a little bit different today. But I want to introduce our guests that we have, Stephen Wojno, which is not spelled Wino at all.

   But I want to introduce you because you have an interesting story. And I'm really excited to dive in today because a lot of cool things to cover. At 26, you became a shareholder of a bankrupt company and took it from 7 million to 28 million. You guys had 70 employees exited that business with over 300 forexed. That business in manufacturing, tier one, and then for the next 20 years of your life, you kind of moved over to professional services. We're from Detroit manufacturing. I want to dive into that story.

   And most recently, you've now dove into the world of AI over the past five to 10 years, right? I want to dive into that. You're interested in longevity and health and wellness. And yeah, I think you mentioned to me that you're going to live to 105. And I think a lot of entrepreneurs are interested in. I mean, a lot of my podcasts are about personal family business. It's not just business, but Steven, thanks so much for coming on. Thank you for having me today. And a little caveat is selfishly, so you recently just got a phenomenal new position at a company.

   So you're chief intelligence officer at self consulting, which is pretty awesome because to me, the big thing is it's AI. And what does that even mean? I think in my ecosystem of entrepreneurs, every visionary, every CEO, anybody and everybody that I talk to, everything is about AI. What can I do? What can I, you know, blah, blah, blah. But what I find is, you know, you and I have talked immensely about like the EOS model, all these different types of entrepreneurial frameworks, but something interesting is going on because those platforms are all about the CEO, the visionary needs to delegate and elevate.

   But I see a lot of visionary stuck with AI and trying to figure these things out. So I kind of just want to dive in with you on where you add in this whole AI journey and what is your kind of gut reaction with everything that's happening right now? Yeah, great question.

8. 3:29
   Stephen Wojno

   So let me put AI to the normal layperson, very simple. Until now, very few people know what it feels like to delegate to total competency. If you manage great people or lead great teams, you know how it feels to put someone in charge who will get it done, get it done right, and get it done without drama. That's the kind of delegation is pure joy. Therefore, when we follow the US model is to delegate to elevate. What we find in AI today is the ability to offload some of the monotonous, transactional workload that has to get done and provide the opportunity for what you hired somebody specifically to do.

   What their specific unique ability is or skill set is to actually provide value to your customers, enhance the relationship with them, and do it in a timely manner. So it's kind of an assistant that I use it today. So where have I evolved over the last five years of working with AI was specifically, as these models have evolved, I have evolved as well. In the beginning, you just have to start being an explorer with the model, give it a try, see what it can do. You're not going to reach utopia on day one.

   I'm still learning myself, but I've gone from a user to now able to automate workflows for myself. I've been able to now get into what we call the orchestration mode, where now I'm able to step away from my computer and work will be getting done by my agents, which I have built a workflow to accomplish goals and get my daily tasks done, do project management, enter information through APIs, through other systems that are key clicks to me are timing and navigating through a user interface of a specific model and in the small business model that in entrepreneurs that we work with, they've got about five or six siloed systems that they have to enter data into.

   Over the last 10 years, I've been working totally in business intelligence. So we try to remove all those clicks and turn that time into strategy and customer facing.

9. 6:01
   Wes Mathews

   So if I understand correctly, just to kind of regurgitate a couple things. One, it's like if there's simply call it simple tasks in the business, something repetitive, that's the lowest hanging fruit that this AI can come in and take over, right? Correct. That's what I hear. So if I'm an entrepreneur, I'm a business owner, I think what I want to really dive in today is like, what's the entry point? Because even I struggle, I've had a pretty successful first company, I'm a pretty good visionary, I'm not the smartest guy, but I struggle with, man, where do you even jump into this?

   Because you hear about it, you're like, I just want this to take over and do all this great stuff. But as a visionary, somebody's got to do the work. But let me back up even before that, because I think it's important to understand what is a model, right? And what is an LLM? So you referred to models several times in what you just said, but can you break down what a model actually is? Yeah, so let's first start, talk about what an AI model is. It's an LLM, which is a large language model.

10. 7:08
    Stephen Wojno

    It's really a text to text conversation that you have with a model that's been trained on everything that's out on the internet. And each of them have specific strengths. I have a subscription to all of them.

11. 7:25
    Wes Mathews

    And all of that being like, are we talking like ChatGPT, Gemini, Grok? Are there any others that are important? Yeah, Claude,

12. 7:32
    Stephen Wojno

    Claude, through anthropics. Those are the major XAI. So there's about five major players. And you'll hear about them in the news every day. But basically, the ones that I use are specifically Claude number one, because of where its position is today, Gemini number two, because of where it came in the last three months. And you'll see me say last three months, because this stuff is ever changing. It's like a horse race out there. Each leader is pulling forward and heading in a different direction. And then I do have ChatGPT with open AI, which again, just this week came out now with like a health app.

    So that's interesting, but not for us in business. Claude has been the one that has gone after the enterprise business level, and especially a small business community that, as I've told you before, we can now bring the technology of a 200 person company into a 20 person company. Therefore, we can we can now get the insights brought to us at such a fraction of a cost. So that's a large language model, then within your business, you'll have a financial model, you'll have a delivery model. And all those models have to for AI to be successful, let's kind of take stuff away from AI and you ask what the first step is in a company.

    And it's actually built in the US model as well as you should be documenting 100% of your procedure. So let's talk about the hard thing about documentation, SS procedures in that in the past. I call it a blank slate syndrome that we've had in the past, when you give an employee the task to document a process, they don't kind of know where to start. So I like to use AI, the large language model as draft one, is it going to be perfect? It is not, but we're going to go ahead and iterate off of that.

    And it's going to evolve into a completed documented process that we're all going to follow. That's the hardest thing in an organization is AI can only be effective in offloading off of your human workforce is if we all agree that there's one way to do a process, we can always change it, we can always refine it through evolution, iteration, whatever your terminology is. But we need standard operating procedures to be able to be documented first. Then when I talk about I go to orchestration is if I can take a standard operating procedure, I can offload that to a mini agent, because you're hearing all these words out there, but it's just a workflow.

    So let's say we're going to get a certain one task done. And you also said in the beginning, oh, we're going to bring AI in, it's going to solve our problems today. It is not, you have to go ahead and have it solve a simple problem. You have to understand how to communicate to it. And once you do understand how it operates, and you're going to start training that model, just like that model was trained on something much larger than your company. This is where you're going to train it on your business.

    Therefore, when you take a simple task, then you're going to go ahead and get that automate. And then it's what we say is we put that onto the flywheel. So that's offloaded off of human. Up next, let's do the next task. Let's do the next task. And then what you're hearing about agents is where you string several tasks, completions or workflows together, where you start the process. And once the first task is done, the agent now knows to go get the second tax, then they know how to go get the third task and that it's complete.

13. 11:12
    Wes Mathews

    So it's interesting, right? Because like, I think a lot of my personal experience and just like my colleagues, people I know, it's, yeah, I've used ChatGPT, right? And I use it a lot for email correction. If somebody requests any documentation from me, like the other day, somebody said, Hey, can you send me your bio? And I want to know this and the other. And for me as a visionary entrepreneur, like if you ask me to sit down and write something, I could stare at a blank screen all day, and I'm not going to get anything done.

    You know, ChatGPT is what I use, because that's all I really know, does a great job of understanding who I am. And it kind of, it gives me information and it kind of gives me the training wheels to work off of. And that's where I thrive. Like if I have something of substance, I then can take it to the next level. But then I feel like a lot of visionaries are probably stuck there. I don't think it's just me alone. But now we start talking agents, I'm like getting stuff done.

    And I'm like, okay, that's what I'm missing. Like I don't, I'm still pulling the levers. And I'm still arguing or debating with my AI, right? That knows me because I've vetted so much information. But what's the first step? I'm like, Hey, like, how do you instead of me going to my email and right click copying and pasting, like, can you do that for me? Like, can the AI actually just go in and actually do all this stuff? Like, what does an agent actually mean? And how do you start to think about that or implement that in the business?

14. 12:37
    Stephen Wojno

    So great point. And you're no different than anybody else. And you're no different than me, let's say a year ago, because I put a lot more time and effort into this discipline. And the models are ready. Let me give another stat. That was a white paper that was published in q4 of 2025 is that let me be clear the AI technology that we can subscribe to today, we'll do 47% of our workload, you just need to know its power, and how to harness that and put that into the right direction.

    So let's talk about your agent strategy is how do you go about doing that? So yes, first, we're all just going to go ahead and interact. And it's kind of a chat bot, we call it. It's a personal assistant helps you understand maybe something that you're not disciplined in. Maybe you have a question that needs to go to your attorney. But you just kind of go ahead and ask ChatGPT, or a large language model that question up front. That's great. I'm not telling you to get rid of your professional services, when you're really ready to make a decision.

    But as you said, it gives you great insight that you otherwise wouldn't know. Let me give you a live example. I was working at a client yesterday. And we're talking about loss carried forwards for financial statements at year end. And I'd sold one of my last businesses prior to 2018. But I am not a CPA by trade. But I can call my CPA get this answer. But we just said, okay, let's ask ChatGPT what the answer to this is. Because we were worried about how long the loss carry forward could be written off for.

    And in the old days, when I owed companies old days, 10 years ago, it was only three years. Today, now I learned that it's indefinite. So that was a good move because we're working on an M&A strategy. As we get closer to the deal making, we will go ahead with that insight that we gathered from a 30 second conversation with a large language model. It now gives us the understanding of how to how to think and structure the deal so that we can finish our financial model to do the virtual acquisition that we're looking after.

    But without that, I would have had to call some money. So you're right, it's a nice personal assistant that you can feel like asking it a question that you're not the sole disciplinary on that. Therefore, what we do is I ask it a lot of questions. It's made me more intelligent. But let me be clear, I'm not going to eliminate any of the other things when I need an attorney when I need an account to do real work, then that's when we deploy them. So for the small business owner, that's great value.

    Because when you have a thought or a question, you just need that answer. When it comes to an agent is, and your second scenario there is, let's say you're working with a Google sheet. The technology is out there right now that you can through English language, describe what you want to be done in a Google sheet in Excel, or PowerPoint word, you just name the document that you want to create. And through English language and iteration of your own English language, it will create that actual document for you.

    So let's say I have a financial model, and I want to put a forecast in it for the next 90 days. I have the ability to now speak to that LOM, tell it to put these dollar amounts in these respective months. And by training it previously, in doing this skill, it'll actually put those in those fields. And I don't even have to open up the spreadsheet, I don't have to enter them in there, close it, save it, look at it. It's done for me, and I can start trusting it.

    Now let's start talking about how do you get to agency. Agency is, you have to test your agent from the very beginning, you cannot trust it to do everything automatically. It's just not going to do that. That's what you're hearing out in this AI world. That's what's fearing a lot of people that they can be replaced, that this will just solve everything. It is intelligence. It just does not know your business, it does not know how you do things. And until you train it for your specific workflow, workload, company, the way you do things, it cannot do it by itself.

15. 17:03
    Wes Mathews

    That's interesting. And I also want to comment, my own struggle with utilizing ChatGPT is sometimes it's giving me bad information that I can see right away. So that poses a challenge, because if I'm asking it something relative to like a CPA issue of, hey, the depreciation advertising is three years, we're looking long term. And if it says, yeah, it's actually indefinite. And I'm like, oh, great. But the reality is, if that's not true, like I can call it out on things I'm aware of, like, where are we in that stage of like, hey, there's still gaps in the holes.

    But my assumption is, at some point, that's going to be buttoned up where the humans going to make more errors of the AI. But right now, I feel it's flipped. You're correct. So that was 2025.

16. 17:46
    Stephen Wojno

    And so my prediction, and every prediction I give you is just what I see when I'm exposed to on a daily basis is that you got this blank screen with no user guide. Therefore, it will hallucinate, it'll make you feel good. And you think it's giving you the right answer all the time. Well, that's just not the fact. And market has driven us to introduce new technology within those LLMs. I'll give a specific example why I'm a anthropic Claude fan is Claude in September of 2025, introduced what is called the skill.

    The way I like to look at this skill is I am now going to start putting into a document that's controlled that will harness this LLM. And what I mean, my harness is all the intelligence that I get about this specific skill. And in your case, you were talking about amortization, and maybe different accounting terms and stuff like that. So you would build the skill these days. And you would tell that skill to go over to the AI ACP website or some data set that would not give you a bad answer that you could trust.

    Therefore, once you tell it where to go, where to get the exact information, what definitions are, what calculations are, and you start defining that when you now speak to this, it will actually when you want to start talking amortization, it will now have an amortization section in it, to the way that you know is correct and already built, and then it just gives you the information faster. So anthropic did a great job, which is also what I'm excited about. And the industry is that since anthropic came out with skills in September, it had a great market presence.

    Therefore, it will be brought over to Chantew P.T. and Gemini and other major models. So I like that we're getting the architecture that no matter what model you choose, we do have the ability to put the guardrails on it, and really deep dive into the discipline in the domain. So let me go back to your question. When you are a domain expert, like me, I'll talk about this. I've been working with Power BI for 10 years since Power BI came out. I had to learn two programming languages, which are DAX and Imcode.

    Unfortunately, today for me, I feel a little deflated is that I became an expert in that over 10 years. However, Chantew P.T. or any of the large language models will actually write that code for me today. I actually don't write that code anymore, because it's non-value added to me to do that. However, because I'm a domain expert in that code, I do know when it's given me a bad answer. So as I like to say is, stay within your lane. When you're the domain expert, you still need to check it to make sure in this.

    In the evolution of this, I think 2026 with skills and other technology that's going to come out, it's going to tighten the hallucination or the bad answers. By 2027, that's where I don't know that we're going to be full agents in 2026. 2025 was supposed to be about agents. It just didn't transpire. We're just now getting in the agency mode, but I'm also thinking it's later part of this year because we have to build the foundation first. You have to train your own model to do this.

    And how you do that is through a skill or anything else that you can make sure it knows exactly how or the best database to go to get that. You just don't let it go wild.

17. 21:30
    Wes Mathews

    Let's just kind of stick on the EOS model of in a perfect environment, you have a COO visionary, a COO integrator. You've got sales, you've got marketing, you got operations, finance, HR, IT, etc. To me, all companies look like that, that are non-enterprise. There's very defined roles and responsibilities, typically within those. So who's responsible for this, right? And in an organization, because to me, it's the wild, wild West, you've got people that are, excuse me, just doing their day-to-day, operating off their processes, procedures, just doing what they do.

    And now there's this thing out there called AI or language model in ChatGPT. But I'm gonna throw a lot of bullet points, but who's policing it? Who's tracking this stuff? Who's setting the standard in an organization? Who's driving that forward? Because I look at, is AI really just a new way to solve a business challenge where the business challenge might be solved through traditional software technology? Is AI just the new stack on, right? Because what we're doing is my hypothesis is nobody owns it. And it's a problem.

    That's why you are the Chief Intelligence Officer, and we're kind of creating this new thing, which is, hey, now we have somebody that owns it, that can look out for us, one for us as stealth, but number two for client work. So what do businesses do? I know I just kind of just dumped a lot, but that's the big challenge I see of just this kind of mess we're in of, you know, this is bigger than the internet.

18. 23:19
    Stephen Wojno

    Let me answer your first question, who owns it? That's the problem. Nobody knows where it goes. It's not going to sit in your IT department. Let me go back to my domain expertise example that I gave previously is, each of those departments that you named marketing, sales, ops, finance, COO, CEO, those are all domain expertise in their specific thing, but nobody owns everything. And then that's the differentiation and what we're doing here at stealth. And that's what my business background is. I started in the finance department, then I became a business owner, then I became a technological expert.

    Then I've studied intelligence and analytics for the last 30 years of my life. So I now know every single discipline, top to bottom in an organization as an owner of myself, now I get to go into companies and say, I might not be the domain expert in a department, but I can tell you how that department runs. I can tell you how it processes information. And what we do is we're here to support every single department in there. So we will set out as chief intelligence officer, a goal for each department to participate in building each of their respective department silos using AI.

    Then the key to this is when we started talking about agency is, how do you get the multidisciplinary hand from one department to the next department without any inefficiency brought in between. So that's where we understand that it can't be owned by one single person. If it were going to be somebody, it could be a COO. However, the COO is a swamp today because they've got five or six different departments underneath them. So it's like them having an assistant to an assistant to get this done.

    So that's the ownership side of how I feel it should be structured, but you need to invite somebody into your organization. We feel we have the skill set here to offer at stealth is who knows the business process from top to bottom, operationally, financially, customer relationship level, employee morale level, as well as shareholder value. Because those are all three of my customers is I am here to provide better value at a lower cost faster. And to do that successfully, we're going to have to walk first and then we can run and then we can put agents on it of the monotonous, non-strategical thinking stuff that we do.

    In a small business, it's hard to hire all those individuals. So we come in and we would assess the situation, see how we could structure your organization to do that. As we are starting here with ourselves here at stealth is where can we pick the low fruit?

19. 26:24
    Wes Mathews

    And then where do we go to multi department agency to get our work done? And you mentioned something that I want to bring back which really sticks out of my mind as an entrepreneur, as a visionary. I don't know exactly what you said, maybe you can repeat it. But the new 200 person company, we could be a 200 operating person company with 20 people. And that resonates with me because I've been in that scenario with 100 plus employees. And the old me was the business challenges, throw bodies at it, like hire people in.

    That really bugged me, right? Like I exited the company, you have an exit. And one of my things was I don't want a lot of employees. And I hear that in the entrepreneurial community, like employees is always the Achilles heel. There's the A players, B players, C players, D players. But if now there's this world of like, yeah, now you can invest in 20 A players. Because to me, AI isn't to replace, it's to refine and invest in your A players and say, hey, like, let's make your life easier or block and tackle.

    And now you have this technology that can really make you go faster, stronger, etc. There's a lot to that, though, because now you have that one person within the organization, right? And if they leave, there's no redundancy. Or is it right? Or is it you're feeding an LLM or an agent model where it's actually kind of easier? So like, what are your thoughts on that? Because I think that's an interesting way to think about this. Hey, we're not here to replace it, it's to amplify. Great, great questions.

20. 27:56
    Stephen Wojno

    And it's twofold. So let's talk about growth and scalability. Number one is with AI. Yes, we have the ability to run a 200 person company with 20 individuals with this technology as their sidekick. What we need to understand clearly is we're going to work with the employee to amplify them in their skill set. Therefore, that's number one. To be able to do that is, again, we feel that there's 10 to 20% of efficiency productivity gain just by using AI as a tool and a partner, assistant, whatever your word is.

    Then what happens after that is once you get that efficiency there is now what you need to build second, which was your fear is you have 20 people. Sorry, let me go back to your A player versus that I might be able to hire I can hire a B player today, and give them AI and make them an A player. And I can actually get a C player up to a B player. Because as you know, with me, operating businesses, as well, I used to own 300 person company, and I'm not interested in that anymore myself either.

    But now I have the ability to take a C player, put them up to B player, put a B player up to an A, and then the A just skyrockets, right? So therefore, that's the part where that's first. Second is your knowledge base. What we do through structuring this workload and agents, everybody is going to perform the process basically the same way. Then you're going to document the knowledge level. That's what's missing in small businesses today. You have great people, but it's what I call is it's a people driven company, not a system driven company run by great people.

    Let me be very clear, people are the key to making this work. We're just going to institute an intelligence system into that, so that if you have one person leaving, there is no gaping hole anymore. We have built the intelligence into the system to pick up where that person left off to the best of our abilities. And then let me talk about on hiring on the way in. Today, I treat AI as my intern. I am teaching it everything that I do about my business. When it becomes qualified to become equivalent as an employee, I now can hire an intern and give them my AI system as a partner, which now elevates them faster.

    Then I just won't need as many employees in the future. So as your financial model, your delivery model might be for every four customers, I need one of these specific roles. Our goal is in the next 18 months to say for every one of these, I need five or six customers they can handle. That's where you get scalability with less human capital, but you haven't lost the intelligence. And intelligence is the key word. Your humans do have experiences and intelligence to bring to the table. You're just bringing what they call artificial intelligence to be their assistant to help them make better decisions with a broader set.

    What brings me to the table today as CIO, I just got this 40 years of experience of overall business, success and failure. And I've been through a lot. The question is, how do you replicate that? We're going to build that into a system so that even I can be replaced someday, because it's doing it from not from a replacement standpoint, but if something were to happen for succession planning, or any other negative effect to an organization, remember, it's all about the company first, then it's about the employees and its customers.

21. 32:17
    Wes Mathews

    Let's use this today. I use ChatGPT, and I shove a lot of information in there. I have the paid version. A lot of companies let their employees use the free version. Some do the paid version. A vision for me would be like, if I'm not a landscape company, but let's just assume we are here for a minute, or we have some clients, hey, we've done this commercial building at 50,000 square feet, or sales is talking to this prospective client, hey, I've got XYZ hospital, they've got 50,000 square feet, they want to completely gut their landscape job, blah, right?

    My vision would be AI should understand, maybe go to Google Maps, look at it, get all the dimensions, look at 10 other variations of hospitals, take a look at the community, and then spit out what the company's done in the past, and give a comparable quote, and do all that stuff, right? Because if you're a new employee, or a new salesperson, have we even done hospitals? Because the old way of doing businesses, who do I talk to in the organization, right? The sales manager, take their time away and say, hey, I've got a hospital, they want a new project, let's spend all this time.

    But to me, it's like, that's where AI can really come in and create efficiencies. But where do you start to build that infrastructure? Because the reason why I say that is, when's the best time to plant a tree, right? 20 years ago, or start to plant it today? A lot of people need to realize, how do you start doing this today, so you're that much further ahead in a year? Because if you're just waiting on the sideline, you're going to get cooked. Yes.

22. 33:59
    Stephen Wojno

    If you haven't started already, I'm going to recommend that you must start in 2026, because we're not turning back. It's only getting smarter and faster delivery point of intelligence. Let's use our landscaping, because there's a great agent right there that I could produce a quote or proposal that would probably, I would visit the client, I would talk to them, I would take some information, I'd go back to my office, I talked to my staff, and then maybe I get back to my prospect, maybe a couple days later.

    How would you like to get back to your prospect in like three minutes, while you're sitting right there with them? Everything you described right there. You did a great job. Thank you. Let's start with Google Maps. We can hit an API to Google geofencing. We can go ahead and list the hospitals in a radius, state, whatever you want, how a normal thing looks like. Again, let's just bring it, you're building a model. You're building a model of what a hospital looks like. You're building a model of what a 50,000 square foot building looks like as a model.

    Number two is you have your operational data. You have your old quoting information which sits in your database in an Azure client somewhere up in AWS, somewhere we can get a hold of all your past proposals and actual projects that you've actually produced so that we can have a proposal versus actual cost. We can tag that data as well. Then you have your financial model that says how we want to price things going forward. All three of those components can be brought together, put into an agent.

    When you speak to this, and I guess I will talk about speaking next, but when you type into this agent today that I'm sitting here in front of this client, they got a 50,000 square foot building. It's hospitals. It normally looks like E, B, or C. You can rate these things however you want to do it. Give me an estimate that I can give them right here. It'll produce it in a matter of 30 seconds if I were to guess to do that. It'll do its number crunching if it takes two minutes.

    We're getting a little bit impatient today. I get to laugh because some of the people that I do things with is they're like, wow, it's taking 90 seconds. I'm like, are you serious? This used to take hours. We're going to get spoiled by that solution that you provided right there. Let me do a little segue into that. Everything today is still keystrokes. In 2026, it's been out there for a couple of years now at voice, but voice is going to become a major player in 2026, I feel, because it's refining its model.

    Your speech is two or three times faster than how you can type. If we think these LLM models are performing at a superior performance today, wait till we start throwing speech at them and they can get 100% comprehension of that. It's just going to elevate the speed of this as well.

23. 37:12
    Wes Mathews

    Going back to that landscape model, if you're an A player, you're in sales or whatever, to be able to give all the right inputs is where the value is. I'm valuable because I can look at every scenario and know every stone that needs to be unturned to say, hey, here's the three players that are making the decision. Maybe the AI agent can go scan Sally Sue and understand that she's got four kids and she lives in Northfield, Michigan, and she likes long walks on the beat.

    There's a lot of data that could go in and also say, look, okay, and also my sales manager, we have to make a 20% margin on this. Give me all the proposals, all the close won, all the close loss, run some analysis around why we think we lost it. Give me the market of arborvities versus this and the cost. There's so many different things and that's where I feel on the sales side is going to create so much value and efficiency because everybody at the end of the day that owns that 50,000 square foot building, they're going to say, what do you think this is going to cost?

    Sales like, wow, I'll do a good job. What's your budget? They can say, hey, we've been in business for 25 years. I can run some analysis to show you that on a per square footage per basis based on what you want, here's the low, the mid, the high, and it's accurate data to do that like that. The old way of doing this, let's say AI didn't exist, that's a huge project because now you have, okay, I'm going to build something like this internally, which could cost millions and millions of dollars, or you hope and pray out there in the market that some software company has thought of this and they call it tree landscaping software.com and you're going to go into there and they're going to have 50% of what you want.

    You're going to sign up for it. You're going to pay for it. Somebody internally, like your IT department or somebody is going to have to go in there. They have to build staff around it and you're like heavily invested in this thing or you custom build it internally. How does AI change this? Because I look at it of the millions of dollars that would have cost me in my mind, like if I wanted to do this,

24. 39:17
    Stephen Wojno

    I have a feeling it's a lot more efficient, a lot more affordable and it's actually doable without a massive headache. It is and it's going to become less expensive going forward because remember all those systems, and again, I'm a programmer for the last 20 years so I can tell you specifically what I had to learn, the skill set to actually perform all that, and the cost that it takes to do that from a human capital standpoint. That's what maybe what you're hearing on the marketplace is anthropic cloud code in 2025.

    What was the number one AI success for 2025 is it's a great coding agent because it rewired the whole software engineering program. What you can license, I licensed cloud for $100 a month and I never hit my limits yet and I use it every single day. For $100 a month, what we used to have to pay thousands of dollars a week for to get a dedicated software engineer to build it and test it and do something for us internally is now, okay, it's $100 a month plus my time, but once I build it, it's done.

    So I heard on a podcast just coming in this morning is that the future is you would go to software, but software is going to start coming to you because I feel that each individual person we're going to be able to develop software that makes them tick. What I mean by that is that a lot of people like to use software a lot of different ways and software has defined us to go into that software and put constraints on us. Now, us data guys, we like to go through the back door through SQL and everything else.

    We don't like the UI, but we understand the strengths of the UI to get the information into the system, but today's systems are just a collaboration of APIs, but we can call that API in seconds. We can pass information to it. So we're going to start designing custom software, maybe just let's say even here at Stealth, we might have custom software and how Wes operates. We might have custom software and how Steve operates because we have unique abilities on how we look at information, how we want to interact with systems, and my attention to detail versus your attention to detail might be different.

    Or let's go back to the sales department because you hit on two things I wanted to hit there as. In our proposal, we forgot all about the research component of LLMs. So when I need to go in and see a company and get research about that company, I'm not going to go to Google anymore. I'm not going to go look at all this archaic data about this company. I'm going to go and ask a question to AI, and I'm going to turn on the deep research button, and it's going to give me a full understanding of the company, how long it's been in business, and anything that it can find out about that, including full staff that works there, drilling down to LinkedIn, who you need to understand specific.

    So sales is all about relationships first, and then we just can productize our proposal process to that. But to build a relationship with somebody, this is LinkedIn 10x, we call it, where the LLM will go out and get information, anything that you need about a person and or a business. So that's very key to salespeople to do that. Now, the way to process that is, how do we get a proposal in their hands, and then salespeople putting information to the CRM is a necessity that we need in the business, but it's never enjoyable by a salesperson.

25. 43:04
    Wes Mathews

    They never do it properly anyway. Yes.

26. 43:07
    Stephen Wojno

    So what we just asked the salesperson to do is to just speak to this box or type into this box, what they just landed in English language, and it will arrive in the CRM system for us, the way that we want it. So we're going to train the model to key it in the way we want it, not the way that the salesperson is going to key it out. We just need it to speak a certain language, and we're going to train that. And that's easier to train for your people, as well as your salesperson says, Oh, wow, that's all I have to do to get this into the system.

    And I can go on to my next sale, and I can go do my next research for my next proposal, because they're permission-based. They're hunters and gatherers. They're not data-enters.

27. 43:50
    Wes Mathews

    No, and I can tell you from my experience, number one, I default into the sales marketing bucket, you know, I'm visionary, but streaming down to the core, I'm all about relationships and, you know, every client engagement I've ever had when we start to look at their data, their CRM, and I've worked with companies under a million, I've worked with companies over a billion. The billions are kind of like, they're a little bit more efficient, but not really. Everybody's data is a disaster. And the pain point always goes to sales just won't do X.

    And I think salespeople are what they are. And it's like, been a lot of meetings where like salespeople get bashed, and I kind of protect it, because that's just like my default, but I say, do you want salespeople putting in the right data or just bringing in the deal flow? Like that's the hard part. The data is the easy part. So I think what you're talking about is a pretty awesome solution to the future, because that's just the part that, you know, it's not the salesperson's highest, best use of their time.

    Their highest, best use of their time is building the relationship. And it's all the other stuff we talked about that can be offloaded and delegated to systems, agents, et cetera, or just speaking into a box. Holy cow.

28. 45:02
    Stephen Wojno

    And let's get back to that salesperson and or any person, because we're doing it here within stealth in the first month of my arrival here is that we're going to each and every employee and asking them, tell me your top five dislikes about the portion of your job. It is unpleasurable. I will then take that and meet you halfway that if I can alleviate that pain point for you with a better system, I don't know that I can eliminate it. But if I can alleviate 75 to 90% of that workload for you, and you just got to give me a little effort to get it, we're going to use AI to be that seamless integration, because you're right, you cannot change people.

    Okay, you want to change people, but you're not going to change people and behaviors 100%. We have biases, we have the ways that we like to do things. It's just that technology has the ability to meet you halfway now. Therefore, the employee will see the benefit in this that oh my gosh, I don't have to do that anymore. If I just say 123, and I put it in this way, it's done. And I can go focus on the things that make me money or spark my interest and keep them engaged.

29. 46:27
    Wes Mathews

    Yeah, even like reminders to write like the old boring way if you're a salesperson, you're working on a HubSpot, heaven forbid Salesforce or some type of pipeline platform. It's like, you know, what's the follow up? You know, when should I connect them with the like, that's the tedious stuff that just slows people down, right? It's been analyze my pipeline, my contacts, what is some low hanging fruit? What are some opportunities that can be generated this week relative, you know, there's just so many things. I think that's where a players, that's the differentiator.

    It's like, I have the inputs, like, I'm strategic. My failure, my D work is sitting me in front of HubSpot, like, I prefer not even to have a login to be 100% honest with you. Correct. I like the platform just as just to your point. Hey, I'm talking to Stephen Wojno. He does this like he's in my pipeline. He's a prospect. I think that there's XYZ opportunity. You know, based on the research you find, let me know the best way to follow up. Maybe he likes long walks on the beach.

    Maybe he likes golf. Maybe he likes this, like come up with some concepts and some ideas to help me maintain. That's where I see this like the people that can differentiate from the crowd. That's where I see this just amplifying, you know, getting them to the next level. So with that being said, right, so if you're a business, where do you start putting this information or taking all of our Google file? It's interesting. Like I look out at our company today, my first company, we had all these phones and all this crazy wire leading into this data room.

    And we had servers, we had all kinds of stuff today. I have a little Comcast box that's like this big that powers the internet. And we use Google Drive, right? Yeah, you're in the cloud. Like cost nothing. It's like a couple hundred. I don't even know. Like, you could trash my computer and I'll be up in Apple in two minutes. Number one, they probably drive it to me in four minutes, but we don't need any of this stuff. Well, with that being said, wherever people's data is today, how do how do you start to take that data, pull it into what we're talking about to start this, because we're hiring six people this quarter.

    You know, other companies are hiring or they have data and they want to share in that data. Where do you put it or do you just leave it and connect this to it?

30. 48:44
    Stephen Wojno

    Well, where do you even begin? So let me go back. Your system of record will not go away. So let's talk about HubSpot. There is no desire and no anticipation from what I see today for HubSpot to go away. It's a great system. It serves you well. The question is just being able to get into HubSpot. And as a new employee, for example, I'm not a HubSpot expert. It'll take me a long time to get up to speed on HubSpot. However, if I build this AI-LLM interface to HubSpot that I'm just speaking to HubSpot, let's go through that question that you said is you have this podcast with Stephen Wojno today is how would you like to have a agent that knows that you're going to have this call this morning to be able to put a brief in your inbox in the morning all about Stephen Wojno with whatever's in HubSpot, with whatever you tell it to go ahead and do.

    So therefore, getting into a computer today, and this is what I've worked on with business intelligence and data analytics for the last 10 years is people still rely on going into the computers, filtering reports, exporting, copying and pasting and doing all that stuff. That's where I've been the last 10 years is we just use through Power BI, data analytics, Tableau, data lakes, or whatever, we set up the data foundation that you can get dashboards to do all that stuff for you so that you're not playing around with your data because that's a very time sensitive.

    And we don't want to do that anymore. I don't even really want you my vision of the computer of the future is it's that box that you see or it's that microphone you have in front of you. You don't need to we're going to build the wiring to bring that that's what I mean, we're going to bring the computer to you. We're going to bring that computer to you on what the questions that West likes to answer. And it's going to train it. And that's what these co pilots are known as is, hey, I like my information sliced and diced this way.

    Steve might like it a different way. But this is the way West consumes it, and I can understand it. Therefore, when I asked for how's our sales trending for this year, what's my pipeline look like? When should I return a phone call? When was the last connection point I had with the client? You set all those standards. Today, I'm doing it in cloud skills. And then the skills will go to all the platforms. So that's the technology that we're using today is we're defining a model there.

    That's your business model. That's your operational model in your sales department. Here's what we expect when something goes into our pipeline, we expect that every three to five days we follow up on something or we're doing something to keep that lead live and or we're kicking it out of our system and saying it's not a good lead because that's where a lot of sales pipelines are technically stale and or they're real, not they're not qualified leads. So our goal in the business is to get all those unqualified leads out of there or close.

    You can don't get them out once you put them in. But the question is, then we get a lot of data analytics off of that information. I'm managing that pipeline a lot better. But I don't want somebody going in and trying to find my record, trying to find what I do, trying to do this and that. If you can ask a question about me, it can give you that information in a matter of minutes. So we'd rather bring information to you instead of you going through a system to get information.

    It's stored in your database, your cloud environment today. I do not care whether it's a word, a PDF, whether you use Google products, whether you use a Microsoft products, whether you use a notepad. As long as we have access to it and all the elements have access to everything on your workspace, in your cloud, everything has access to your laptop. I just go ahead. I'll forgive where I have a file on my system. We used to search like file explorer drives the heck out of me.

    It really can't find my file. I've got it in five different folders actually, which is another issue, but I can never find something when I want to find it. Now I just go and ask the LM to go find it for me and miraculously it finds. It's 2026.

31. 53:01
    Wes Mathews

    When do you think in your crystal ball that a humanoid can prepare my meals and fold the laundry in my home?

32. 53:12
    Stephen Wojno

    That's my second interest outside of business and longevity is I'm really excited about humanoid. And I'll probably have one of the first ones with Susan comes out because you'd have hit that discussion. So actually in the fall of this year will be the first humanoid put into a home. It's already out there in the markets. You can put down a $500 deposit and get this to come there. For those that don't really know, the humanoid market is a huge, massive market out there. They're working in manufacturing facilities like Amazon, Tesla, and all those out there.

    Basically, the reason they can't put that into your home yet is they don't know how to control it in an uncontrolled environment, but they're working on that. So think of like we're talking about AI and LLMs with the server farm. They have homeowner farms out there that they just have a whole subdivision of homes that they have engineers working with these robots in every single home. And every single home is built differently. So just think of all the testing that's going on right now. I apologize.

    I'm very weird. I follow this stuff, but I think it's like Jetsons. Jetsons is here. Therefore, I'm already going to tell you, I saw a product that you can actually buy that will cook your meal for you once you put the ingredients in it. And it was really intriguing. While you're asking that question, everybody's out at CES 2026. It's one of my dreams that I'd love to go to, except for I don't know that I'd have a checkbook after I come back because all the gadgetry and all the innovation is happening.

    Is that going on right now, right? Yeah, it's in Las Vegas this week. And every year this week, it's amazing. However, so to answer your question, I will predict that there'll be humanoids in households in the next three years. I know I will have one, but I'm an evangelist. I'm an explorer. I'm curious. I'm not fearful. If you have fear, drop your fear by using it, adapting to it, understanding it, and use it to your benefits. Yes, they're always in every single industry technology that we've built in America.

    There's good people and there's bad people. Let's just focus on the good things that this is going to produce for people. The value that we're going to be able to provide to our clients, the value I just saw that AI is now coming out as a wellness program for employees to take some of the stress level off of them personally and professionally of the demands of the world on them, which I think is a great play. But all you have to do is you have to work with it.

    You can't believe what you're hearing. It is amazing. I started my journey five years ago, just dabbling in it to seeing what it's all about. Today, my life has changed. I'm a better person personally. I'm healthier. I also am better to my customers because I can provide more value at less cost and faster too, though, and it's more accurate. The better that we can become and let this be an evolution and set up scarcity, it will provide abundance to this world. And in the health market, I'm super excited on my goal to be 105.

    I really now think since I've got half of my life there to go, I want to make sure that we either come up with something or cure cancer, diabetes, or anything else that might be thrown at me in my later years.

33. 57:04
    Wes Mathews

    I now see that that can be eradicated. I feel like I probably needed therapy, but I use my ChatGPT where it's like my release valve, where it's not really judging me. It's just kind of black and white and answering questions. And for me, I think what it's done, if I reflect back on my first company, I used to tie things to revenue per head and what we've been able to do with this company and the short amount of time we've had in like two and a half years.

    I mean, we just close out our second full year. I'm over double revenue per head than where I was in the first business. I think a lot of that has to do with just me having the outlet of my LLM to be able to get my ideas and gut check and just I spend a lot of time, hopefully in a healthy amount of time. I don't know. It's a lot, but I spent probably the most amount of time on ChatGPT, but it's really gave me the right training wheels to do things that I wouldn't otherwise do.

    Let me talk about that because there's two things that you hit in there. In 2025,

34. 58:08
    Stephen Wojno

    the number one use of ChatGPT was for personal psychology reasons, coaching, mentoring. It's easier for us as individuals to actually ask a computer to be harsh on us versus a human. Our emotions are hurt. What a human says that to me or we put up some defenses, but when you ask for honest, credible feedback from LLM and you really give it the words to go ahead and hit it. Don't sugarcoat it. Yeah, don't sugarcoat it. You're in your own private space. It's between you and the computer.

    Do you adapt to that a lot easier that if you and I were to get into a room and we were to talk about our differences and our assumptions that we've made on our conversations, it's totally different. That's the positive that I look for that. Let's talk about revenue per employee. Yes, if you've done a double of what you expected in one year, intelligence business unit that we're going to set up here, I expect that we want to in the next year when we end 2026, we're going to actually cut that.

    We're going to actually go from you doubling that, so I'm going to say we're going to double it again. However, you keep on asking the same question. What are we going to do to guarantee that? We have to build the foundation for it to eventually run a lot more on its own. It cannot run on its own. You cannot turn it on today and just expect it to work. You have to start today and in one year to three years, and as technology involves, you will see your return on investment.

    You'll see short-term return on investment. Your low-hanging fruit is very minimal. You and I met already this week. There's two things that I'm going to help you guys out with. We're going to have them finished by today of a little bit, a couple of things that you guys asked. It's kind of nuisance. I don't want to do this anymore.

35. 1:00:14
    Wes Mathews

    Most things add up. The nuisance add up. They take brain power. They take your mind away from what's important. I want to go back real quick on the personal side, because as I'm thinking, I'm like, you know what? The one thing, and I want to get your thoughts on this from, because you're more to me like the scientific analysis behind the scenes, but I also feel like in the last year, by utilizing the LLM though, I've probably cut my relationships or me taking some of that information and asking a peer.

    I come from an entrepreneur organization, EO, so it's a tribe of other entrepreneurs that are all kind of doing a business and living life that all come together for a talk, but I feel like in the past year, I've really amplified where I've been able to solve or get out a lot of my whatever with a machine, which is now probably dropped my communication with people I know or in the past, I might have had those discussions where now it's not happening. That to me is kind of interesting in the sense of what happens,

36. 1:01:27
    Stephen Wojno

    because I can't be the only one that's experiencing this. The reason that it's probably dropping today is because we're in the early evolution of this technology, so we all have to spend time to learn it and understand how to use it. I believe that once we, let's say you've used it for a year or so now, you're good at stuff, so you don't have to spend as much time, now you can get back into those relationships. Yes, there's a lot of people that say they don't want relationships anymore, that's very negative.

    That's not good for anybody, because again, the world is built off of human relationships and connections, and especially as you get older in life, the companionship of a person versus a computer are totally different. I get drawn into my curiosity that I always want to be excellent at things, so that's why I'm continually learning with this, but then I know when to turn it off to do that. In my business life, when we're in the ring, we're going to use it to its best capabilities for everybody, and I get the joy of seeing people use it for the first time, get that aha moment, and say, wow, what about this next?

    That's the exciting part of my job, is once I can teach you how to fish, you're going to fish by yourself eventually, and basically, you're never going to turn back, and you're going to say, what did we do with that? I can tell you today, what do we do without this? I write better code today than I wrote 12 months ago, because I don't write it too much anymore, but I'm a domain expert in it, so I know about it. We're going to give you superpowers to become even better person than your actual domain knowledge is all about.

    Not that we have to know everything, and I want to be the dumbest guy in the room, but when asked a question, I want to add my insight, and I want that insight to be grounded in good information, because great data on time, you can make great decisions with it. Bad data is where we're going to make a shitty decision, it's going to cost our company, it's going to upset a client, and that's what my mission is. How do we use truthful data, unbiased data, that's the other thing that as we sat over this whole podcast is, when you ask AI to give you the utmost truth about your data and benchmark against others, it will give you that.

    You just have to be ready for it, but it's not a person giving it to you, so somehow our bodies and our minds are able to adapt to that a little bit easier. Maybe we're in private, it's just between me and my computer, but now I'm able to do something about it, and then with my personal life, I have a whoop, I have a aura ring, I have an Apple watch, I track all my biometrics, I get my blood tested twice a year, so just like we do for a company, I've got to keep this body alive for a lot of years.

    So what am I doing regarding how I sleep, how I exercise, and how I feed my mind with information in this crazy world to keep my own sanity? I'm not special about this stuff, it's hard for all of us out there. No, that's great, there's a lot.

37. 1:04:56
    Wes Mathews

    I think that we'll have to do a podcast part two on the things that are already interesting and having discussions with you on that, because it's not just about business, there's family, personal, other stuff that, to me, it's like people, it's like, well, business is business, personal is personal, no. When you're an entrepreneur, all this bleeds together. Personal family business, to me, is a big tie in and what are you doing in all of those buckets. All good things must come to an end, we're like an hour ten into this podcast, it's been super great, kind of diving in here.

    I think that, you know, we're going to follow up this year as stealth here locally, where we'll have an event and kind of dive deeper into kind of maybe some of the takeaways and how to's for AI. But in the meantime, if, you know, somebody has a massive or simple business challenge, and they're interested, they heard something here, your name's Stephen Wojno, it's-T-E-P-H-E-N-W-O-J-N-O. But how can people get in contact with you if they want to learn more about what we just kind of discussed for the past hour?

    You can get in touch with me.

38. 1:05:57
    Stephen Wojno

    My LinkedIn is updated here. I'm at Stealth Consulting, proud to be here as of January 1, excited about what we're building here. And we'd love to have an opportunity to talk to you. Also, my email address is [email removed]. Please feel free to reach out to me in those two venues at this point in time to just start a dialogue and a conversation. I'm an educator. I like to inform you on what availability of information is out there for you to advance and understand how to get where I was five years ago very fast.

    And all I can say is what took me five years will take you six months now because the technology is really there. But if you start today and you go on a journey and you keep on doing it like any other habit in your life, it will become a normal process for you.

39. 1:06:57
    Wes Mathews

    Awesome, man. Well, hey, I appreciate the insight. I think there's a lot of value today for anybody that's listening. So I appreciate it. I'm looking forward to a recap to see how much has changed from this conversation until the time we jump back on a podcast. But really appreciate the time. Thanks so much, Steven.

40. 1:07:12
    Stephen Wojno

    Hey, great. Thank you for having me.

41. 1:07:14
    Wes Mathews

    This has been another episode of Entrepreneur Intel. Thank you for joining us. For show notes, for other episodes, please visit us at entrepreneurintel.com. Until next time.