What 20 Years of AI Taught This MSP

[00:00:00] Shadow AI and Shadow IT going all over the place. And that sucks. And cyber people are like, lock it down, lock it down, lock it down, lock it down. No matter how much you lock it down, people will use a personal device and they'll try to do things to circumvent your controls. And then you're working against the team. So you're creating a natural fiction set. You're saying I don't trust them to adhere. I get all that. We're zero trust as well. But my philosophy is, is if you build a golden road, which is the Autobahn, people will choose that because it's the best option.

[00:00:27] So giving them something to use, it turns off the Shadow AI and Shadow IT. So you're actually giving them a road to go down. Hey guys, Damien Stevens, host of MSP Mindset, founder and CEO of Servocity. Today is my blessing to interview John Snyder, and he is the chief AI officer at Nexogen.

[00:00:55] Now, John is one of the most fascinating people I've met in the area of AI. He started in 2003, decades before we all heard of ChatGPT. And he's worked with some of the largest companies in the world in their AI implementations. Yet, he still operates in MSP. He still thinks about the literacy and the governance, more importantly, the culture within the MSP.

[00:01:21] And I've never met anybody that could take the decades of wisdom and the hundreds and hundreds and hundreds of reps and distill that down and how to I could apply this to my MSP. So if you want to gain John's wisdom, don't miss out on my conversation with him today. John, thanks for being on MSP Mindset. Thanks for having me. I'm super excited to be here.

[00:01:45] Yeah, I'm so excited for this conversation because I don't even know how to introduce you with all the companies and all the things and how, like when people say they were early to AI, they mean like 2024 or late 23, right? You were much earlier in that way, but also you have got this really fascinating community angle.

[00:02:08] So, yeah, help us understand like why you're doing these things, your background a little. So the why, start with why or the why, the purpose statement. I always get, I love that. I love books about it. I explore it, the neuroscience of it, all that kind of stuff. I just love it. So that's just a great way. So my why is a journey. It's not a one thing or that thing or all that kind of stuff.

[00:02:33] And first off, I started off really weird in business in general to create Nexogen, which is my MSP. I was at Best Buy working on the tech bench prior to it being a geek squad. And people would bring their computers in and they always wanted us selling on the tech bench. So we would learn how to sell the basket like a complete solution. You got to have CDs and your modem.

[00:02:58] You have to have the right graphics card, battery backups, PSP, performance service plan. So you learn how to sell warranty, what the values were, a lot of transactions. And then people would come in and just with broken stuff all the time. And their computers were just ridden with just crap. You realize like humans are not good at computers naturally. And for some reason, I was always kind of good at it. And I met my business partner there. His name is JJ Schaefer. And we were always super interested in excess.

[00:03:29] Meaning the computing power was like, this is, it's almost like prophetic what my business partner said. But I was reading, the singularity was near back then as well. So I'm a big reading nerd. Oh, wow. And that was sort of like, that guy had calculated like the amount of compute dictates the amount of intelligence. And then when you couple those two together around this time, you'll end up with this like sort of better AI version.

[00:03:54] And we were on a whiteboard one day just trying to think of how to get excess capacity out of processors. Because typically when you're using a, typically a user uses a processor, they're using like 3% of the total processor. And virtualization didn't exist back in 2002. So we were like, if you could put multiple operating systems on one piece of hardware, you could get way more efficiency out of that processor. And he wrote on the board, whoever has the most processors at the end of the day wins.

[00:04:23] And we then developed Nexogen as being a competitor to VMware, which we created our own virtualization suite, which was really crappy. And him and I basically just came together with like $500 each. And on the $500 each, you weren't able to compete with like VMware, which raised like hundreds of millions of dollars.

[00:04:47] So we would have a really cool system of like patched together technology to get people basically a utility computing environment is what we called it at the time. And that allowed us to charge customers for processors, storage, memory. And we were the first, one of the first in Cincinnati. And then at that same time, we were exploring other technologies in college and so forth. So my business partner's background is computer engineering. I was a systems guy.

[00:05:16] We both had played with neural nets pre-2003. So we had explored that concept. We had, they were absolutely horrible. They were horrible. I mean, they were super weird. They didn't have enough data. We didn't know what we were doing. You just kind of like installed it and then like put data at it and you build it. And then you would run things at it and it would not function coherently. And you'd be like, okay, that's not really interesting.

[00:05:43] But we would try to get it to do weird stuff. And it just like throw an image classification with something people tried to go with at the beginning with. So that's kind of the early like part of AI, I would say, is like right that early, like 2003 of curiosity. Not a professional, not in a research lab. But like, yeah, we were playing with it and like knew it existed.

[00:06:06] And then as we started Nexogen, we basically had the hosting side, which was occurring with the utility computing. Then we hosted websites and designed websites. And then we had the help desk and infrastructure and server sales, which just came from customers saying, hey, I don't understand what renting a piece of a server or paying utility is with you. Sell me a server and a backup and a tape backup and then come once a month and pick up the tape. That was what they wanted.

[00:06:34] So we sold them that because that generated income and we went back in time. And then have since then come back to where we started at, which was more futuristic approach. Now it's going beyond where we started. But we then are now about 90 people. We have 24-7 help desk, 24-7 security operations center. It could run it in its own company. We do IR, forensics, all that kind of stuff.

[00:06:59] And then a network operations division that's also 24-7 that does patching, firmware, network, all that kind of stuff. A cloud solutions group, which is independent. And then our AI research lab we've added on top of all of that, plus the consulting practice. So it's like a lot of capabilities in one building. But our customers leverage it, a lot of them in unison. So they'll use 3, 4, 5, 6, 7, 8, 9, 10 different services.

[00:07:28] And you end up, we're also at Microsoft CSP. So you end up with just like a lot of different SKU management, billing, these kind of things for customers. And you naturally just grow into where customers bring you in. And we've been very organic. We're not PE-backed. We've never taken on money.

[00:07:45] And the different kind of thing we've done, which is kind of, I'd say, a little bit different than typical MSP, is we've started 15 or so businesses out of Nexogen, which are now standalone businesses that are doing very well. One of them being in Smart City, which was kind of an AI-backed company. We have a friend who started a payroll company we backed, which is now a SaaS platform plus a payroll company. That now employs like 80 or more people. We've got real estate. We've got an events company and these kind of things.

[00:08:14] So with all of those as a leader of that and working with fellow leaders, when 2022 November came around, that was the genesis for me on the new generation of AI, which was ChatGPT was released, version 3. I had already played with version 2. And I had played with tools that were leveraging version 2, which have already been like you can't even find them anymore. But they were big for like Jasper AI or something like that for marketing was big. And then it just like disappeared.

[00:08:45] So we were playing with those. And when 3 came out, like absolutely mind blown and realized where that was going to take us. So the first thing I did is work on myself. I freaking prayed. I freaking was like, will my kids be okay in 30 years? Because this thing is going to change so much. Is there going to be political situations or all these kind of like things that are coming on today?

[00:09:10] I was just like really brainstorming and sort of talking to my group about it and becoming more vocal about it on social media, big on LinkedIn. And then once I went back to... Let me break that down a little. Yeah. So you were really crazy early with neural nets and creating your own VMware competitor, which I love both of those. And then when like most people never heard of chat GPD 2 or even 3. So you were still quite early there.

[00:09:40] But I want to go back to something that I think is pretty important because you said first I worked on myself. Yeah. And you prayed and you worked with a team. Tell me about why. So I look at myself as being responsible to the families that I support. So we have around 230 families that we support around my community. And we base everything we do every Monday morning at my company. It's private. It's what we do. Nexogen at least. Not all of them. And it's not forceful.

[00:10:08] We say we start with yourself or God. Then family. Then work in that order without hesitation or friction. But we will not force this on you. But we know if we make an investment in our people, our people will then show up stronger for our customers. So by investing in our people, we are investing in our customers. And that exact statement came from the head of my network operations department, Douglas Hammond. And I put it as his quote for the team to hear every week. And he loves it.

[00:10:38] Which is a great thing. Culture is built by the team. So you're just kind of like creating the space. And then people are working within it. And then for me, though, I have to show up my best self. And I have to make decisions that could impact their livelihoods. They also have kids that are going to go to college. They also want to get raises. They also have career growth opportunities. They also get depressed. They have their own personal illness problems and things like these that you have to work through.

[00:11:08] So I take all that very seriously. So yeah, I had to understand that technology very well at the beginning. Yep. So my process for understanding it was I went back and I found the top guy in Cincinnati. His name is Kelly Cohen. And he has basically worked on what's called genetic fuzzy algorithms, which is a part of what's called explainable AI, which is basically taking the black box away and creating it where you can see inside of it.

[00:11:35] And he had worked on projects such as the military defense system in Israel, which automatically targets missiles and shoots them out of the sky. So I then started working with him as a research partner and taking classes from him, auditing classes from him, and then doing grants together with him and his team. So Kelly is a mentor of mine. And then a bunch of other people like that I now work with on a regular basis.

[00:12:05] So that was then from there, I had the confidence to, you know, in 2023 to start making some moves and some of the organizations. Hey, guys, let me let you in on what's new. Just for the month of May, we are doing weekly Agentec AI build sessions. Now, this is for you, whether you're technical or not. You can come knowing nothing about it. And an hour later, you're going to ship something real in your business.

[00:12:34] Creates real value and there's no lock in. It's yours to keep. So if you want to go from I'm wondering what to do, I'm overwhelmed, what's my strategy, what can this stuff even do to really getting most of the videos answered? And more importantly, no PowerPoints, no talking, just building, then make sure to sign up using the link below. I want to talk about that because you say the confidence, right?

[00:13:00] Was that when the idea or concept of kind of an AI lab was born? Because I feel like most of us as an MSP, we're still trying to figure out what we do with it, period. Never mind, like you very firmly plant the AI lab as a related but kind of separate piece. It is. It was. So you have. Okay. So I look at. I say weird stuff, but like one of the things I think about is what's called temporal intelligence.

[00:13:27] So it's the speed at which things work dictates how we can coherently interact with things. So like the plants behind me, they work at a very slow speed and they don't talk and they don't have higher level cognitive cognitive capabilities. AI systems are sort of like intelligence. I don't even know if I'd say there are yet. There's some. It's kind of it definitely acts like it. And yeah, there's a speed at which they work.

[00:13:53] So when I'm interacting with this technology to kind of get my lab set up, it was the farther you go out in the future, kind of the faster things are occurring. So it starts at research papers. So people are exploring concepts and they're putting out research papers. That is usually two to three years ahead of what you're seeing and sometimes five years ahead of what you're seeing today.

[00:14:22] So a research paper will come out first years in advance. Then they will take that research and implement it into a functional working part of a product. And then you will get the benefit. That's how EDR works. That's how a lot of these like typical like older AI systems that we've worked with classification and so forth. They all work on that. The difference between ChatGPT is it's more of a universal approximator.

[00:14:48] So it's trying to do everything for everybody, which is more general because it has all the different diverse data sets. And then that's just slightly different. So our research lab started writing papers on explainable AI and neural net technology. And we've published them on novel or new things that don't exist yet that we've figured out that are not in the market today.

[00:15:14] Which then allowed me to understand after also purviewing research across the market, what will be coming and what the things that will come will look like. So it informs the rest of my practice. So we go out very front. Here's what's happening here. Then we know which vendors are going to be picking up what because we can see what the signal should look like. Oh, yes. Gaussian is a big thing right now people are working on. People don't talk about that word a lot.

[00:15:42] It's a mixture of extroverts are a big thing people are working on, different groupings of them. How those signals work, how we can kind of see what is going to be coming, especially with like self-improvement recurring loops. That then I know where it's going to be in like three or four years or have some visual. And I wanted that forecasting capability because then I can make decisions for my company.

[00:16:05] And then I can make decisions for my customers and I can advise them on like solid technology versus hype-y technology, which you were kind of alluding to in pre-talkings. How does that turn into the kind of enablement and adjacent to community work? What would you say there? So self-God, family, work, then we actually say community and proximity. So we work together in the office.

[00:16:34] We come in as community in the office and we spend money the closer you are to us. So if you're a supplier and we meet you in person, that means more to us than a supplier we do not. And if we have, we can help someone in our community versus outside of our community, we will weight that in our decision-making. So proximity matters and that all drives human connection, which I think is going to be more important as we move forward.

[00:16:58] So that's first, that's Nexogen's kind of the complete Monday morning meeting kind of like statement on top of the other piece. However, that's how I live my life because I can't affect things that are far away from me. I can't think about them all the time. If I do, it's wasting my time. I can get to work with things I can reach out and touch. So my event series, which we started at first, we're like, hey, this is going to be very big. We went to MIT.

[00:17:26] We went to a big conference there. People were spending money like crazy there. I mean, Damian, they were freaking spending trillions of dollars, like hundreds of billions were being spent in a room that I was in, not a big room. Like I'm talking 2023-ish, early 2024, which lit a fire under my butt because those people were moving. They were not messing around. And what they were trying to do and the language they were using was like, we're going to eliminate industry.

[00:17:54] And I'm like, when I started asking questions, though, technical, like, can you explain why this occurs? Can you figure out why hallucinations are happening? Can you figure out how to stop hallucinations? How are you stopping jailbreaks? How are you reducing your risk exposure? All these kind of things. They did not have an answer, nor did they care. And the reason they didn't care is because they don't, the people in that research functionality piece, they do not look at cyber. They look at functionality.

[00:18:25] They're what I would deem like manic functionalists and optimists. So they didn't look at the risk side. They were just looking at like, how do we keep this thing moving fast? Which realized to me, there's a gap in their thinking and an opening in the marketplace around responsible AI and around, the more you can explain something almost from a safety standpoint, the more functional you can make it. So even though they might be faster, there's a second mover advantage to understand it better and actually create an improvement.

[00:18:55] And a lot of times, Midwest or early inside the country markets, not the East Coast and West Coast, which have all the money, we adopt their technology because they were first to market and they have all the money. Whereas we have just as good as brains across the country. So could we create an activation pattern was the thought process in our hometown, so closest to us, and see if we could get a little mini bomb let off to like explosion of ideas.

[00:19:23] So we created an event called Cincy AI Week, which was a community event. Number one is you don't have to pay $15,000 to send next to the coffee stand if you're a sponsor. You're on stage sharing your ideas and sponsors are a part of the community. They're not dominating the conversation. We made tickets 100% free first year.

[00:19:47] So there was, we fed everybody, tickets were free, around 1,200 people showed up, which is a crazy big conference we were not prepared for. I think we had maybe like three volunteers and like my entire team at Nexogen was coming down running cameras and stuff. I mean, it was incredibly like put together quickly. However, we had 150 speakers or something like this and we got it done.

[00:20:12] And people that were on stage were on stage for the amount of knowledge they had and we changed the event structure. So we don't do formal PowerPoints and like walk me through bullets. We did one minute long and three minute long startup sessions. We called them lightning rounds. We took our talks. We reduced the length of the talk. So we said, hey, you have an hour long talk. Make that 20 minutes. Now let's do three talks in 20 minutes or panels.

[00:20:42] And it fit the people's attention span for that group, which like more information faster. So we're able to fit 125 to 150 speakers in two and a half days, roughly. Two days, really. And people then get a ton of information in and they stretch and grow. So our feedback is I can't even process everything I just heard. However, in a week later, they'll email us back and be like, wow, I learned a ton. Because it's so much, it's almost like a force feeding of information.

[00:21:12] So once that kicked off in Cincinnati and we're like, that's great. We then launched a whole company around this, which is called Enterprise Technology Association, which now runs conferences in Nashville, Atlanta, Cincinnati, Columbus, Cleveland area, Toledo. And now we're working with the White House at AI Congress will be this month, later this month with the head of the Labor Department will be there. So we've now trained last year, 2025. First year, we were kind of like getting everything set up.

[00:21:39] Last year, we trained, so to speak, around 15,000 people in person. And that's a separate organization from Nexogen, which born out of it, that direct idea of let's make impact across our country. Because America needs to win this thing. I'm a big fan. I grew up here. I love it here. I love our freedom. I freaking love our people. I mean, I love our country. I like hiking it, swimming in all the lakes and everything. We have amazing space.

[00:22:08] I want to make sure we did that the right way. Healthy, add some voice to that for the next generation, leave it better than you found it and all that good stuff. So that's part of that process. So how do you connect this super early? Like you're informing things because of your lab and then you've got this amazing community that you've built that's now growing to so many cities and so many parts of the nation where you're doing in person.

[00:22:38] And then how does that connect back to or inform your MSP? Like how does that? Are there just separate businesses or is there a connection? Separate business. Two other people, co-founders, run ETA. So I have tremendous belief in other people that I partner with. They run that thing. I invested in it. I help them anywhere I can. Shoot, I'll fix printers during the event. I will wear a volunteer shirt and cheer on the volunteers new training that day. Like I'm very much a utility player there.

[00:23:08] And I will get on stage and speak. And I try not to be a dominant voice because it's not about me. It's about the community. And that's one thing that we made sure of in the beginning is we actually take a very low profile. A lot of groups come up there on the conference. You'll only see their face. You will barely see me at my conference, except I will hand you a coffee when you need it. And I will get you to where you need to go from a VIP standpoint and all that kind of stuff. So that group runs independently. The feedback loop is all the innovators are coming to these events.

[00:23:36] So we found in Cincinnati, there's like people that worked at Google Brain. There were people that worked at Meta Labs. There are people, all these geeks came out to this conference because this is their rock show. So those people speak. I learned from them. You're getting community with these top AI leaders. They're informing you of even more. So you're broadening. You're not only getting clarity. You're getting a more broadened clarity. You're seeing hundreds of use cases since 2024, like hundreds every month. There's just so many crazy use cases out there.

[00:24:06] That just informs you. And then that helps you then understand your customer. Plus, there's executive roundtables we do. So you can hear other CEOs and other presidents and other business leaders, IT directors, things like that. Their problems, their failures, and all that kind of stuff. Like we've had things we've done internally where we've had failures on. So all of that does connect. Built a community that we now participate in and it's helping us. So I guess that's kind of how it's worked. So I kind of sponsor it and help both sides.

[00:24:34] But I'm typically day-to-day, I'm working at Nexogen. I'm working on cyber problems and the typical MSSP duties as a CEO, working on relationships, new sales, that kind of stuff. Not many MSPs start an event business that trains thousands of people in a year. Not too many build like an AI lab that gets to see the things years ahead. What should other MSPs, what would be helpful for them to know?

[00:25:03] Because if most people aren't in your shoes, what would it be most helpful to understand like right now? I don't know. I'm not saying like Nexogen's not necessarily. My path is not the bright path for everybody. Like I'm a bit, I'm always, people always tell me that I'm a cocaine squirrel. I do not take drugs. I, my brain just works like that. So it's constantly active and it's very scattered. Like I think about a lot of things. I can connect it together and create weird connections. That's just me.

[00:25:33] So I'm kind of just being me in this world. Okay. So if, when I'm talking to other people, I'm like, take your time. And honestly, I do as well. We don't, we didn't jump and just like put things in. We're very like change management means a lot. But we do not add, we try to simplify, reduce the amount of complexity, but I want to stay informed so that I know what shelf we put in. So if we, if we, I have this concept, I think of a shelves.

[00:25:59] So if we put something in place, that shelf, we're going to lean on that thing. Like think of your ITSM package or like ConnectWise or Halo or whatever ERP kind of thing you're using. And you don't want to change that out every year. Like that thing is a shelf and it's got more books that are getting put on it, more data. And that's your business. So like that, that is, and it gets harder to manage as you add pieces. Like we had data lakes and all these other things we put in there and add complexity.

[00:26:28] But so to me, I'm saying with AI, I keep it very simple. I have a lot of concepts to work out, but I usually keep it pretty simple. And I focus on use case, not tools. And what is a use case? A use case is a coherent explanation of a problem that can be solved with automation. So what does that mean? That means to me, I kind of have a cheat sheet here.

[00:26:57] One is, is we have, we've defined a problem. So that, that exists. We have a, we have a, Hey, we have a problem. Okay. We think this problem can be solved by data. Okay. Can it, and do those two things even connect? And can the person that thinks they've solved the problem with the data explain why the data actually makes sense to solve the problem? If they can do that, then you have more coherence to the use case.

[00:27:23] Now, do we have domain expertise around the problem we're trying to solve? Which means I can now bring in professional experts inside my four walls to then add to the problem statement, plus the data, plus the expert. If you compress those things together with someone who knows how to do automation, you now can create a repetitive task. And you can do it in an effective way. Now, that to me is simplifying it into very basic bread and butter.

[00:27:52] Second thing is, is how do I measure that that was successful? So that's very simple KPIs. I'm going to say here. Number one is what is my accuracy score? An accuracy score means I actually understand when the thing gets something wrong and what right looks like. Most people do not. And in coding, that's very hard to get into an automated measurement standpoint. So you have the ability to generate a lot.

[00:28:20] The more you generate, the harder it is to understand if it was correct or the accuracy. And you have to be very careful with judgment, all this kind of stuff. Second one is, is what does my false positive rate look like? I'm sorry, false negative rate look like. So false positive, false negative, I'm probably going to confuse this statement. The key point is, is when the system actually says it's accurate, when it's not. Yeah. You need to have a metric that tracks that. So one is the accuracy.

[00:28:49] Second one is the false accuracy. False positive. Okay. The third thing that you can do on a KPI is the latency, which is how fast the thing takes. And then the fourth KPI would be your ROI. So can I get a financial CFO to actually develop a KPI, which shows me what my usage is, my costs are from a utility standpoint, because now we have token consumption, all that.

[00:29:14] Plus the problem statement resolution and the labor saved or the problem solved. I can go through use cases where I can show and track the KPI that says, I am now saving $10 million a year or something like that, because this is running with those three criteria are defining success. And then I can show the financial success. If that all four of those are together, then I consider that a completed project. Like, like that's actually running the correct way. Now we're just like, that's an engine.

[00:29:44] And then now we're just changing the oil and like doing small upgrades and being very careful when we do because new models come in, it changes things and screws up the balance of everything. So what is your take on like, you've been doing this for a while now and not just AI, but also kind of the, for lack of a better way of saying, like figure out what to do, like how to go to market.

[00:30:06] Obviously everybody will do their own, but I'm kind of curious, like what's your take on like, do you, you know, products and vendors and consulting and like, how do you, I guess the better question is like, what are your thoughts on how you best help people with AI? You know, using AI? Great question. Who am I? Who does it start with? Like, who am I? So who is Nexogen? We're, we're a great Microsoft shop or CSP for 20 years. We, we know that stack inside and out. We know how to secure it.

[00:30:35] We know how to do it responsibly, all that kind of stuff. I kind of didn't want to be the Microsoft guy because I was like, everybody's going to be the Microsoft guy as my thought. And I was like, I want to do something different. So, but, but that's really who we are. I mean, we're mainly Microsoft. We're mainly in the co-pilot space. We'll work with like Anthropic or other models as well. But that's, you know, OpenAI, Microsoft.

[00:31:04] Google, people that work with Google usually work with like a Google focus shop, I found. As far as tools, there are so many you can use. We're trying to leverage ones that are in compliance and so forth. But what's your philosophy on tool first? Use case first? General purpose. Like getting, okay. The question I, I say the question I always come to is like, what if we do nothing?

[00:31:33] So one is there's shadow AI and shadow IT going all over the place. And that sucks. And you're like, cyber people are like, lock it down, lock it down, lock it down, lock it down. No matter how much you lock it down, people will use a personal device. And they'll try to do things to circumvent your controls. And then you're working against the team. So you're creating a natural fiction set. And if they don't, you're saying I don't trust them to adhere. I get all that. We're zero trust as well.

[00:31:57] We're also, but my philosophy is, is if you build a golden road, which is the Autobahn, people will choose that because it's the best option. So giving them something to use, it turns off the shadow AI and shadow IT. So you're actually, you're actually giving them a road to go down, like, like, and you're not just saying no. So I see a lot of organizations.

[00:32:24] The easy button was Copilot and is for Microsoft groups. And the reason why is, is because it adheres to the compliancy standards. Microsoft's taking on some liability. You have to look at legal language. I'm not going to make any promises here. And that shifts. But you want to be very comfortable with your, your legal language or, or get to know it very well. We've done a lot of exploration and it's challenging. And then general purpose though, in my mind needs to come out and be available. People need to be able to create images with these things.

[00:32:53] There is liability in all that. For instance, these models are built on illegally collected materials. So you can produce illegal materials. And I'd like to take it back to where we had websites. And when your website was online and you used an illegal image and you didn't pay for it, you would get a letter from some lawyer who would say, take your image down that we have a picture of that, you know, you didn't use legally.

[00:33:19] And then they would send you a letter saying it would be $2,500 to pay the bill of the bill you didn't pay on the front end. You would pay the bill because then they would come after you with like a $50,000 fine or something like this. That same thing is going to happen in AI. So governance, training, employees, enablement is part of that process. AI literacy, they call it. Responsible ethical usage. Because even though today it's a wild west, that stuff will come back to haunt us.

[00:33:49] And if you built a shelf on that, they're going to come knocking. And that's not a reason to not do anything. That's a reason to get more informed so that you can do it better and more responsibly on the front end. And then when you build is more valuable later because someone else didn't do it that way. So when people are shopping for tools or whatever to buy, if you have that, they're going to be like, yes, you did it the right way. And you picked a responsible group to go up.

[00:34:13] So the way I did that for tooling and to get really in with groups that were going to explode, like Anthropic and so forth, very early, was I actually looked at the AI Alliance. Have you seen this? AI Alliance is a group that formed in 2024 or late 2023 between like, I've never seen companies move this fast, by the way. Meta, Google, Microsoft. Yeah.

[00:34:43] I think X was in there. A bunch of behemoths all came together with the AI Alliance to like share information quickly. What was interesting though was, is the startups that also joined. So when I saw the startups that were accepted into that group, I called all the CEOs. One of those, two of those companies have already been purchased like and sold for like a billion dollars plus.

[00:35:10] But I became first partners in North America with them. And, but they got bought so fast. It's like, wow, that was crazy. So I remember talking to a guy in Europe and he's like in an apartment with a mullet and a mustache in his kitchen table that probably cost me 20 bucks at a thrift store saying, how did you get ahold of me? And I told him and he's like, that's crazy. Here's how our product works. I knew about it. He's like, not many people know how our product works. We're only selling to certain people. He's like, yeah, I'll sign an agreement with you. I signed and agreed with him.

[00:35:41] Six months later, I have a call with him. He's in his London office with 175 people at a glass table. No mullet anymore. No mustache surrounded by suits as a CTO. And we were talking a different story. So like definitely a weird journey. And then now he's already exited and probably at a beach somewhere. So that guy, I haven't talked to him again recently, but like, that's kind of my journey on that. Trying to find innovative groups that big boys trust is a big thing.

[00:36:08] What are you most interested or excited in right now? What is the like way you can help people or use case or what is it that's got your interest? It's top of mind. Top of mind is right. Like things like I'd say like closer to present is where I try to focus on. But AI enablement and literacy is sort of something I'm constantly like, it's like a river that I'm like wanting to swim in a lot.

[00:36:36] And that's because it has big impact across multiple places. Second piece is training specifically for IT teams. So I look at IT teams inside of organizations as the engine that run. They're the pace at which that organization can move. How fast can you move a project? How fast can you fix a ticket? How fast can you implement a new technology? The IT team dictates all that. A lot of groups have been trying to bypass their IT teams or they talk crap about them and all that kind of stuff.

[00:37:06] I don't like those groups. That's not helpful. IT teams are working their butt off typically. They're typically, no one has the answers they need. People leave positions that had key information. So you have to go figure that out, how it was built in the first place. Sometimes it's two people removed. You're coming into a new job. Expectations are high. Time pressures are, I mean, all that's very stressful. That's why I have a picture of nature behind me. Because if you're in IT, I want you to see some trees more often. Okay.

[00:37:32] Having said that, that heartbeat that they run at, that pacing is extremely fast. And AI can speed it up. And if you teach the IT team about the technology, then they can be in control of how it's deployed and then grows inside of the organization in a responsible way. So we spend a lot of time directly training IT teams. So that's very, very, very important to me. When you do train them, they usually do look at you as a little bit of a leader.

[00:38:02] And they will maybe call you for a tool or there is sales potential there as well. But that's an exchange, not just an ask. You're giving. I'd say that's a good mentality. Yeah. Second one is governance. Focusing on AI technologies that actually protect and govern. There's a book called Uncontrollable that was written by a guy named Darren McKee. That is a great book to read that is easily digestible.

[00:38:29] He goes over very clear examples of where we're going to run into problems when we get to what's called super intelligence. And it's basically on how fast things move. If you put an AI model that can complete 10,000 tasks, if a human, let's say, okay, let's say you have a human in and let's say one time out of a thousand, they make a mistake.

[00:38:58] You're like, I love this person. It's one of my best employees. If someone's doing 10,000 an hour and they're making one mistake per hour, that is now your percentage of accuracy actually has to increase the faster we go. So IT teams need to understand these concepts and then be more rigid in their thinking with implementing new solutions over time. Today, it's more about like you're putting in a basic workflow today.

[00:39:27] It could be in the future that things running incredibly more fast than you put it in today. And we kind of do it quickly and not necessarily – it's so easy to implement. You don't think about all the other steps that could go wrong. So more testing, more governance, these kind of things are very important to me because it's so easy not to. Like you just – and you're like, why should I?

[00:39:53] And thinking long term, Bezos always says it's like a big human weakness. I agree with him that we just don't think about the consequences like five, six, seven, eight, nine, ten years out. It's really hard to. So I really try to put a lot of tension there. I get excited about that. I try to express that, put energy around it, put it out in the world, manifest it. Hopefully, everybody starts thinking that way.

[00:40:14] Then the last thing I would say is that data quality and data management is really going to grow like a lot. Mark Zuckerberg made a key acquisition of around $20 billion for Scale AI, which literally just did data labeling and annotation. I don't know many MSPs that have data labelers and annotators on staff. Probably should.

[00:40:44] Most of the data that goes in a model that you want to be accurate probably needs to be accurate, which means that someone has to go in and make sure they clean the data, put it into a nice format, very consistent and standardized, and then drive business. That's a new business line for you. Go do that. Go spin up a managed service on data labeling and so forth. Your customers need it. I guarantee it. They don't even know they need it, so you've got to educate them. But that kind of thing has also got me very excited.

[00:41:11] The last thing I wanted to get your thoughts on, John, is you have such a broad perspective. You've been thinking about this for a while. How do you lead your team? Because things seem like they're only going to change faster. And I know from my perspective, sometimes they're going too fast. Sometimes they're not, you know, we don't all adopt at the same rate. We don't all feel good about it. Sometimes I'm excited. We save a lot of time.

[00:41:40] And then, you know, I got somebody that came back to my team and they're like, I save six hours a day. But like, do you still need me? Like, so, you know, what's your take on that? Like, how do we lead in the right way? I've come up with this more recently. I wrote something about it on LinkedIn. I didn't really add this, but like distance, spacing and pacing. Spacing and pacing to me means so much. And I don't think people really think about that.

[00:42:08] And distance means a lot too. So if I show up in a testy or pesky mood and I infect my team with that, that reverberates through my culture. So creating distance when I'm in that manner, which I'm not a perfect human, I can tell you right now, I make mistakes all the time. I'm in bad moods. I have all the same crap everybody else does. Not right now. I'm prepared for this meeting. But I create distance during that space, during that time. And then make sure culturally I'm not putting that out there. Okay.

[00:42:37] Secondly, I have to get over my stuff and make sure I force through that. Second one is pacing and spacing. So I'm going to use a sports analogy. When I played soccer, I was captain. I was fortunate to be captain a few times. I played volleyball. I was a setter and a hitter. I was a utility player. I also played defense, good server. So I also kind of was like vocal. You can see I talk like crazy. So I'm giving people commands and like telling people where to move. And I just see very ADHD, see everything. Um, that is annoying to people too much.

[00:43:07] It's actually, it's actually annoying. So controlling my distance there is actually, um, a big pacing though is like, and, and, and, and spacing is like, if you're in the back, right fielder and soccer and you run that person to the front for a forward momentum, you have now taken a defender with that, with your defender. So their forward now has to work on defending your defender. So you've changed their position. You've added complexity.

[00:43:32] And as that person runs up the field, it clears space that is open now in the back of the field. And it's created new space in the front. That's even beyond the forward. The defender runs past it. That helps a team coordinate together. What I see in business is over dribblers. People that take the ball, want to hold it and dribble around everybody and score the goal. That is not how we play do business. Well, you pass the ball. And the way you pass the ball is you make people have proper space around each other.

[00:44:03] You, you actually encourage them to, um, politeness. Look, I actually, we actually talk about this. Like don't invade other people's space unnecessarily. Um, don't over dribble. If you, if, if you think you can get the customer an answer in 30 minutes is actually, is actually best. Is it better to pause, ask three other people, create the visibility from leadership and now give a coordinated, coordinated answer in an hour on an important thing versus 30 minutes on a quick answer. I want, I scored the goal.

[00:44:31] That that's definitely, um, I feel like there's certain people that are good at what I'm talking about. I'm not the best at it. I can talk about it. Well, um, project managers are great at it. I'm by opinion. If I hire a project manager, I'm looking for someone that does that. It keeps everybody informed. And then they sort of make sure they actually have taken stress off of our team. So as we're doing things faster, our account managers have leaned in to do more. Our service managers have leaned in to do more.

[00:45:00] Our, um, our technicians, engineers are answering. Everybody's going so fast. I'm like, I put the project manager in and I say, look, let's work on a cadence, a realistic expectation that can handle from a communication standpoint. And then their job, I give them authority to slow everybody down. And then actually we make less mistakes. We have to go back in time less often and it's more forward momentum and it's running a marathon, not a sprint.

[00:45:28] And you're actually finding your heartbeat for the marathon, not the, uh, cause this is kind of years, decades. So it's not first out the door, although you want that mentality to move, but you then have to listen to your heart rate. Listen to the heart rate around, around you, find your motion, get the people that really find that groove and keep people flowing. And I think that's the way to lead is more about like creating the space, letting things happen naturally, seeing where stress is occurring, and then trying to effectively manage stress with spacing, pacing distance.

[00:45:57] I feel like there's so many questions. I always said before we started, we probably spend like an entire day, uh, I could asking you questions. This has been a gift, John. Thank you so much for being on MSP Mindset. Well, I mean, thank you for having me on. And then also like personal connection wise, um, your relationship with God that you express openly, um, your story and your journey that you talked to me about privately was absolutely one of the most inspiring things I've heard. Um, I am, I, I'm very fortunate that you brought me on.

[00:46:27] Thank you for the time. And I'm honestly, I want to, I want to listen to your show more, learn about more about you. And I'm sorry, I talked so much. I know this was the interview on me, but I really, really, really, it's super inspiring. Thank you. Yeah. Thank you, John. You're so, you're so, you're very inspiring for people that would love to get ahold of you or connect or find you. Um, what's the best way to do that? So LinkedIn and I, right now I have a mental health awareness, um, month, um, badge on my

[00:46:56] LinkedIn as my profile. And as the top there, my wife is works in mental health. I have had my dealings with that demon. Um, and I care a lot about it. It's a practice in our company, but you can look at me up. If I don't, if it's not May, which is mental health awareness month, then I typically have a horse face chest piece with a blindfold, which is a generated image, which I think describes me in a weird way, um, that you can find me John Snyder on LinkedIn primarily. And then from there you can DM me or anything like that.

[00:47:24] Our website's nexogen.com, www.nexogen.com, N-E-X-I-G-E-N. Awesome. Make sure to take John up on that, uh, whether it's May or after May, um, don't miss out on that opportunity and, uh, the gift to do that. Thank you for being here and bringing yourself to this, John. Appreciate it, Ruth. Thank you very much.