AI + Automation = 10X Capacity for Your MSP

[00:00:00] With an MSP nowadays, I think you have to get in the AI game, the true artificial intelligence game, because you will get left behind. I think by the end of this year, if not earlier, maybe early next year, if you're not an AI, I think as an MSP, you're going to struggle. Hey guys, Damien Stevens, host of MSP Mindset, founder and CEO of Servosity. Today, I got the pleasure to interview Brian Luckey.

[00:00:29] He's the CEO of Integris, one of the fastest growing MSPs. He's also a specialist in human-centric AI. And we went deep on how he saved 50,000 man hours using AI so far this year. So if you want to find out how you could do this, don't miss out on my conversation with Brian today. Brian, let's get right into it. I think the vision you've brought must have something to do with it.

[00:00:58] But I want to get right into the meat of this. You had this crazy number of hours you saved in 2025. You actually measure that. So help me unpack that. Like the hours saved. Let's just start right there. Yeah, yeah, sure. So I believe you're talking about all the automations we did. So in 2025, we really focused on being an automation-first organization. And we didn't focus just on service desk or service delivery. We focused on everything across the business.

[00:01:26] So at the end of 2025, we captured over 3 million automation actions. So it means our automations were run more than 3 million times. And that generated over 35,000 hours of FTE capacity, which is equivalent to 218 FTEs. And people ask me, well, what does FTE capacity mean? So how we do automations is when we decide to automate something, we do a time study.

[00:01:54] So we say, how long does a human take to do this? And then we do the automation. And we say, how long does it take to do the automation? And then the difference, right? So if onboarding. Onboarding a user at the service delivery service desk, 20, 30 minutes. Let's call it 30 just to keep it easy. And the automation takes five because there is some human things, checks and balances, human in the loop. So that's a delta of 25 minutes. As an MSP, the only thing we have and sell are time. That's it. That's what we sell.

[00:02:23] We may have services. We may have products. But ultimately, when it comes down to it, we sell time. And so what we look at it is we have a time piggy bank. And every minute to us is a penny. So we put a little time, a little minute penny, a penny in the penny bank. And so for every time we onboard a user for our clients, when we run that automation, we save 25 minutes or 25 pennies, however you want to do it, right? And every time it runs it. And we have telemetry. So we know what users run it. We know how often they run it. We know which people don't use it, right?

[00:02:52] And so we have all these stats and data. And then we put it on a dashboard. So it's visible to anyone across the organization. And they can see by month what automations are run, which will not run, and all those types of good things. So we have an actionable, very feel-good number that isn't like, yeah, we've just saved time. It's right there in the data. And you can go down and see. You could question and say, well, I don't understand what it is. And then we can go back to the time study and we can show you the information. So we have all the receipts, if you will, though, what the kids say nowadays.

[00:03:22] To be able to say, like, this is actually the time we saved and this is how many times we run it. And here's where it's been run and for what customers. So that's what we've done in automations. It's actually worked out really well. And yeah, we saved a bunch of time. Time's great. But that doesn't mean we let go of people. We didn't have layoffs. But what we did is when people moved on or maybe they left the organization or moved to a different role, we didn't have to backfill them. So in 2025, we didn't backfill 31 people. So yes, that's not 218 people, but that's 31 roles.

[00:03:52] And the 218 people, if you work in an MSP, you know there's lots of things you're not doing. And so we've automated a lot of things that we weren't even doing. So we would literally have to hire 218 people to do all the things and fill all the gaps, look down all the dark alleys that I say sometimes on all the things that we're not doing and automate those. And that's what we did.

[00:04:12] And so a real impact, not only on time savings, on FTE capacity, as we call it, not hiring people back, but we also had a 6% increase in billing. So time billing because people are doing more important things for our customers rather than spending time doing manual things. And ultimately an 8% impact on EBITDA. And that's the magic number. That is a number verified by finance, given to our PE firm, given to the people that matter.

[00:04:40] That's provable with all the data underneath it. And when you think about 8%, at our size, 1,100 employees, 8% is not a small number. And so we're proud of it. But at the same time, we haven't stopped. So we had 150 last year, 152. We're already over 200 this year. We're already over almost 50,000 hours saved. We're already over whatever that number is, almost 300 FTE capacity. So we just keep growing and growing and growing and doing more and doing more.

[00:05:08] And also, we go back to some of our automations we ran because we can probably make them a little bit better. We can make changes. So we're really agile in how we do things in our organization. But everything we do is we look at automation first. So let me unpack that. There's a lot to the back of the year. So much there. Yeah, yeah, absolutely. So much. And I know you're the data guy. So this is interesting to me, right? So the first question is how, right?

[00:05:32] And I think you said somewhere amongst the data with all the time savings, something like 125-ish automations. And so when I say how, what I really am getting at is, right? Did you buy something off the shelf that magically did this? Yeah, yeah. Good question. Was there like a row distribution in terms of which ones gave you the hours? Yeah. So we use Roost. So happy to talk about Roost. We're one of their larger customers. It's doing a good partnership. And very common in the MSP industry to use Roost.

[00:06:02] There's a couple others like Atines or, you know, Zofit can do some AI and automation, a couple things. But we really like Roost. It's very capable, very agnostic, meaning we can use it with a lot of different systems and use APIs, as well as tapping to AI with it. So we have two dedicated architects. That's it. One person that's been here a very long time, knows the business really, really well and leaned all the way in at Roost and is kind of our expert.

[00:06:32] And then we happened to steal someone from Roost, one of their professional service engineers. And so, and they worked on MSPs. So we got someone in our industry that knew MSPs well and just had to learn Integrus. So it didn't take much, a lot of time or a lot of training for him to just hit the ground running. So those two own Roost and own the automation platform or an RPA platform is typically how it's referred to. And that's where I would say 80% or more of our automations are run. We have tools that run out of automations.

[00:07:02] Of course, we use Power Automate. Of course, we use PowerShell here and there because it's just there and we haven't gone to take it and replace it with Roost yet. And then core tools that have automations like a ConnectWise and those type of things. But Roost is our main hub, if you will, from automation standpoint. Not everybody's 1,100 employees. Yeah. Two dedicated engineers. Is this something you recommend to others?

[00:07:27] And how does an MSP begin to even get their mind around this and the mindset around like, where do I even start? Yeah, it's not for the faint at heart, for sure. I think it depends on your size. I think if you're a smaller MSP, 10, 20, 30, 50, maybe even 100 employees, Roost may or may not be the tool for you. I think it depends if you have a resource that has the ability to just lean in and you give them a little bit of time and space. It's not something you just pick up.

[00:07:56] It's something you actually have to learn. That's the downside to Roost. There are other tools that are easier, but not as maybe as powerful. And there's other tools that have changed. We bought Roost, gosh, for almost two years now into the partnership. And there are tools that have come along that are a lot easier to use that weren't so easy to use as well. So I think you just have to find the tool that meets you where you're at in your business and what people you have or person and that skill set.

[00:08:24] And some of your tools probably do some of this already. But that's only half the battle. Figuring out the tool is one thing, but figuring out what to automate, I think, is the hardest part. And there's two ways you can think about this. There is, well, just start somewhere. Sometimes that's just the easiest part. If you don't do anything, then you're not going to see any benefit. But if you just start something and then do something else and do something else and focus on it and get it completed and you just continue, momentum is a crazy thing. And it really helps that snowball effect.

[00:08:55] That's one way of doing it. The other way of doing it is if getting to figure out an area you want to focus on. And those two can be combined. But sometimes focusing on service delivery or service desk is the easy one. It's where your most amount of people are. It's probably where the most amount of automations you have in the business. Or that would be the biggest one. But again, you go back to the other way of like, you know what? Finance is the thing I want to focus on because the resource I have is finance focused, right? So you can take kind of two different takes on it.

[00:09:24] But that's how I recommend. Usually I ask a few questions and you can kind of figure out the best direction for an MSP. But, you know, that's where you start and you got to do something. Otherwise, you're not going to get anything done. Yeah, you got to start with something. What's your take on these? They're now becoming blurry, right? Automation used to mean one thing. AI used to mean another. So am I looking for an automation tool, right? We could always script things in our RMM, PSA. Yeah.

[00:09:54] Am I looking for an AI tool? Is the answer yes to all of that? Yes. Yeah. So I've said this a million times. AI is not, to me, in our industry, is not artificial intelligence. It's automation intelligence. Because I don't believe you can be effective at AI if you're not effective at automation. Automation is a really good precursor to a very good AI result.

[00:10:21] Because if you're not automating it, let me back up. With AI, you have to document steps, right? Because you're telling the AI to do things. It's almost like a workflow. And then you can put some symptom analysis into it. And you can go look at records. And you can do some things. But there are steps you have to take in that AI journey. And automation is a great way to do that. Because you've got to do steps, too, right? The automation needs to know the steps it takes. So it almost gives you an easy blueprint to then say,

[00:10:50] now I can, how can I infuse AI into my automation? And so your answer is yes, all of it, right? You need to find something that has all of it. But if you go straight to AI, you're in trouble. And if you think automation alone will get you aware of it, it's going to get you to a point. But with an MSP nowadays, I think you have to get in the AI game, the true artificial intelligence game, because you will get left behind.

[00:11:15] I think by the end of this year, if not earlier, maybe early next year, if you're not in AI, I think as an MSP, you're going to struggle. Because there'll be big MSPs. But there's a lot of small MSPs. When I go to executive roundtables, I was just at a Power Conference last week at Enable, lots of smaller MSPs. I mean, they're doing some really cool things with automation and AI. And they are light. I mean, they're on our heels at some point, because I'm like, yeah, I feel really good about what we've done.

[00:11:44] And then I go to some of these roundtables and see some of the advanced things that some of the smaller MSPs are doing. I'm like, oh, that's such a good idea. Why am I not doing that? But you've got to get in the game, is kind of what I'm saying. But I think you really have to focus on automation first before you get an AI. And then before all that, though, you've got to get data. I think that's what the missing piece that people don't realize is, you can do automation with okay data. It doesn't have to be necessarily clean. You just have to have it.

[00:12:13] AI has got to be really clean. Otherwise, it's hard to convert to the semantic models and really leverage. Otherwise, you talk about hallucinations. You're going to have hallucinations on the data itself because it's not clean. And so I think when I talk to MSPs, our size or even in the midsize, the 300 to 500 employees, that's one of the main areas that they struggle with is data. They don't have a centralized data lake. They think their PSA, like a ConnectWise or Halo, will be good enough. And it's just not. Amazingly, it's not clean data, unfortunately.

[00:12:43] And that's only one piece of the puzzle. You have to have a whole bunch of other systems together if you really want to leverage AI in the way that you can. Yeah. So I'm going to throw this at you. Tell me what your thoughts are, right? But I feel like you had this huge 10x capacity shift and this 10x vision that aligned.

[00:13:06] And my point is, yeah, you saved 35,000 hours last year, but you spend it on, you know, IP and engineering. And, you know, you didn't make layoffs, right? And a lot of MSPs I don't think will have that discipline or that vision. So what are your thoughts on the industry and the people that have like the 10x, that gain the 10x capacity but have the 1x vision? Oh, that's a really good question.

[00:13:38] I think the problem, especially in an MSP, is we get pulled into different directions. And we see all the beautiful lights and the bunnies and the crickets and the squirrels and any other animal you want to call out. And we're like, we want to go to that. We want to go to that. This is good. And what we think is right.

[00:13:59] And I think that's where we really get in a bad place because that's where you do get in the costs and not the savings you want to get. Or even the impact or the benefits you're trying to achieve. So we always start out discussion of when we want to do something, what's the benefit we're trying to achieve? In the end, what does this look like? And when you start with the end in mind and not just like, I have this idea. And then they go, because I have ideas a lot.

[00:14:29] But if you think about the end result and you work backwards, there's a lot of things we just can't. It doesn't make sense. We'll spend too much money, too much time. And yeah, it's really cool. And we think it'll do something. But if we're not somewhat sure, we don't see a path, we're not going to do it. And so that's where we've been successful. It's just been intentional. Like, let's call our shot. And if our shot's wrong, then let's change it and let's move to something different. Luckily, we've been pretty right for the most part.

[00:14:57] And we're pretty right so far this year, too, as the progress we're making on some of the things we're doing. But I think that's where you see the cost increase. Because we have costs, but even with those costs, even with the savings, it's still a major impact to our bottom line. And again, that's just because we've been intentional about what we're doing and focused to make sure we get it done right. And get it done in the way that we believe that will make the biggest impact to the business and our clients. Speaking of that, how did you get started?

[00:15:28] And some of these things take an investment. Yes. It's not as simple as buying a fill-in-the-bender name subscription and monthly sub, right, and move on. So how did you start? And, you know, what kind of support did you need, right? Did you need to give your team or did you need to say, you know, we're going to – we're calling our shot. We hope it works, but it's certainly the first time. Yeah. It's a great question.

[00:15:56] So I came on in August of 2024. So it's only been here just a short year and a half. And the vision that was sold to me and what we wanted to do by our leadership team, I was all in. And so the good news is that there was a vision, right? And so I just had to fill in that vision. My wife calls me the dream maker. So I was trying to be the dream maker for our organization. And so we agreed on some things that we wanted to do, but also agreed that it would take some investment.

[00:16:25] But what I did is at the beginning of – I didn't add a lot of headcounts. Really not a lot in the beginning. Let me prove this out for a few months and let me show you what we can do. And that's what we did. So we proved it out. And then we're like, oh, yeah, let's – we can do more of that. And so we actually had a slow start. The numbers I talk about, yeah, it was over the entire year, but most of it was in the last seven months of the year. Because that first – we had a few four months to figure it out. We also went through a recap at the end of 2024.

[00:16:53] So we didn't get to do a lot in that last quarter. We were busy doing the recap with our new P firm or our new pension fund firm. And so we just kind of got off to a slow start because we wanted to prove this out and make sure it was right. And then we were able – once we were able to prove it right in just a few short months, we were able to get resources and dig that out. So I think that's the hard part is figuring that out and getting the right resources. It's not my first time I've done this.

[00:17:23] I've done this not only in a corporate, but also in an MSP environment. Not to the extent, not as fast. But I think that also helped having that background. And I've been in software development and operations. I've worked in service delivery and sales and marketing and so different parts of the business as well. Not the normal track for a CIO. And so that helped also. But kind of that all was the end result of kind of where we got to and then just proving it out to expand and get the velocity we needed.

[00:17:53] Hmm. So another thing I want to throw at you is I feel like you've changed the shape of work. I use that phrase often. Right? You've kind of dismantled some of the traditional L1, L2, L3. Yeah. You know, you started to do all the upskilling instead of replacing. So talk a little bit about that.

[00:18:16] But also I want to jump into what that may look like down the road because I think it's easy to say those terms like changing the shape of work. But leading that way and getting into the actual culture where it actually affects career trajectory is a very different thing. Yeah, it's all very interesting on kind of what MSPs have done for decades. The traditional MSP model of L1, L2, L3, really it's inefficient.

[00:18:43] A complex issue that comes in from a customer can easily sit with an L1 for 30 minutes before they even discover that they're not able to resolve the issue and then they move desperately. And then that issue can sit with an L2 and sit for another 30 minutes and wait to be picked up and then potentially resolved. And then that can get resolved that goes with an L3, right? And so you have this L1, L2, L3, all this time to wait. And eventually the L3, that was the person that can resolve it.

[00:19:09] And, you know, you're an hour and a half or two hours into this whole process and the customer is not getting the result that they're looking for. And so it's just a lot of time wasted in having a frustrated customer. So what we've done is introduced intelligent routing. And so we're analyzing what I call the DNA of the ticket. So keywords, asset class, ticket information, all those type of things, and then use data to identify the best person to resolve the issue at the right time.

[00:19:36] So we're optimizing our engineer resource allocation for the best client experience. So factors like, I can't talk about all of them, but like things like competency and availability, specializations, you know, those type of things. We can route a ticket to ensure that it's assigned to the best qualified engineer immediately, which ultimately reduces MTTR and cuts down on escalations.

[00:20:01] So let me try to restate that for those of us that are not as smart as you, right? I don't know about that. But what you're saying, right, is it was always L1, then 2, then 3. And now based on this analysis and the DNA of the ticket routing, you may go straight to L3. 100%. And it's counterintuitive to our thoughts because if I'm an MSP owner, like, oh, it's an expensive resource. I can't have all my tickets go there.

[00:20:29] You're spending more money because that resource was the person that's going to resolve it anyways. And you're actually spending more money because you're spending the L1 time, you're spending the L2 time, and then you get to the L3. If you really think about this process, and sometimes it doesn't even go to L3, it can be an L2. But if you start thinking about getting the ticket to the right person at the right time with the right skill with everything you need, there's a lot of DNA. A lot of work. This is like a year's worth of work. So it's not for the fan's heart either.

[00:20:56] But it's an interesting thing because not only have you reduced MTTR, which we have by 60%, and we flipped the first person resolution. So first contact resolution is one thing. You get the first time, whatever. But the first person, so I mean the first person that you actually called or the ticket's running, that's usually at about 30%, 35% in the MSP industry. We're at 70% to 80% on the industry now.

[00:21:24] So we flipped it over on its head to where it needs to be. And so it's introduced a different perspective on how we service clients. And then when you talk about employees, yes, it's a little frustrating at first because it's a difference in change. Because if you think about an engineer, I'm going to call a few service desk engineers out. The traditional model, you take a ticket and you're like, oh, I got that ticket done. I think I'm going to get a drink of water. And they get up and get a drink of water.

[00:21:53] And then they sit back down and like, oh, maybe I'll just see what's on Amazon. What's in my cart? And so they spend 15, 20 minutes just doing life. And then they didn't take a ticket. Well, now they have tickets waiting for them. They're not pulling a ticket in an unassigned queue, right? There's no word dispatcher. Everything's happened behind the scenes for them. And so it's next one up. Unfortunately for them, they have to do more work, but it's better overall for the customer and they're getting used to it.

[00:22:20] But it is making an impact slowly on the culture of not only our clients, we're seeing that on CSAT, but also the culture on our employees on ESAT. And it's taking time and it's going to take more time, but we're really doing that. And where you're seeing the impact is the upskilling of the talent. So you mentioned it as well. So we can identify gaps in our talent more than we ever have before.

[00:22:43] And so we can get them training, even soft skill training, because we're doing sentiment analysis on the tickets on the after and also the CSAT to understand where the gaps are. And so we can directly hook into our learning management system and send an email to the manager and recommend courses based on the output of some of the lack of skills that we're missing. We also can now hire to the skills we need. So traditional you hire like I lost an L1. I just need to hire a new L1. Not all L1s are the same.

[00:23:13] You could get an L1 that's actually an L3. They're just taking a pay cut, so they want a job. Or you could be getting an L1 that's like an L.2, but they just really showed up well one day. And so instead of that, now we just hire for skills. You know what? We're missing Azure and we're missing SQL and we're missing this. Let's hire someone with those. They could have a bunch of others. Let's figure out what that pay scale is and then let's make that offer. So is this a different approach that I think MSPs will have to get to?

[00:23:41] We believe it's the best approach. There's still obviously time to see that in the case. But the overall stats and the data that we're seeing, like I said, MCTR reduction, the first person resolution, a 5% drop in average age of tickets, 44% decrease in tickets that are open over 14 days. 46% decrease in the tickets that have not been updated over the last 24 days. So we've seen some, you know, again, I'm a data guy, right?

[00:24:09] But our data proves out if we're doing the right thing, being intentional so that we know if what we did works rather than spend a bunch of time on something that doesn't work. Yeah, the data really tells the story there. I'm curious, like obviously in our conversation, you're the argument to not replace your people person. Yes. 100%. What do you think sets the tone? Where do you see, like if we fast forward, what's the org chart of the future look like in a few years?

[00:24:38] Is there still an L1 that's going to like spend their career on the phone? Or do you think that goes away? I don't think so. Yeah, I think, listen, two years is crazy to think about. I'm trying to just get through 2026. Right. I was on a panel and they said, what does M&A look like in three years? I'm like, I don't know. I can't even tell you what's happening tomorrow. Right. But definitely in the next few years, I think there are two paths. I think there are still going to be customers that want to talk to people.

[00:25:08] And I think there are customers that will be okay not talking to people. Mm-hmm. So I think you won't be one or the other. I think you'll have to offer both. And the non-human, it will be cheaper. And the human will be more expensive. And I think if you're going to do this the way that you should, your services should be multi-threaded in that fact. And you can choose what path you want to do. It's like choose your adventure. Which one do you want to go? You want to talk to nobody and have it automated and pay less?

[00:25:38] Maybe some mistakes in there, but it's going to be automated. Or do you want to talk to a human? Still going to be mistakes, but it's going to be more expensive. And people are going to have to choose which way. And if you allow them to flip-flop, that's going to be a question. But I think ultimately that's what's... I see that happening in 18 to 24 months. I think that's what we'll end up being is somewhere kind of this choose your adventure or choose your path model. Or we're doing both. Well, I agree.

[00:26:07] Talking years out makes no sense. What are there... You're both saving more hours. You're investing in AI more than most people in automation. You're also running MSP larger and growing faster than most. So what else do you think are going to be the big changes in that 18 or 24 month time frame that is going to really change the shape of the way we do work, right?

[00:26:34] Because already said kind of the traditional L1, L2, L3 may not be the right model. What else do you think are the big kind of seismic shifts that we all need to be aware of? Yeah. I think you're going to start seeing it. We started seeing it a little bit, but you'll really start seeing it on the sales front where it's more automated, more chatbots, more automated information, more automated quotes. Again, that path of choose your adventure, you can do either one. I think you'll see it there.

[00:27:01] I think you'll start seeing it on the delivery of services, automating services, having AI generate a service or an agent automatically based on... Call Vibe coding if you want to, but a set of instructions.

[00:27:19] That's where we want to get to in the next couple months internally, where we not only have a directory of agents that people can pull from, but they can write a few words and a few sentences or prompts and have our system automatically create an agent for them. It may not be perfect, but something they can use and test and play around with. And I think that's where you can get to from a client standpoint of, I'd like to see this. I'd like to see this data and it just generates it automatically.

[00:27:50] So insights into your information and insights into your network. I mean, that might be more than two years out, but I think you're going to start to see more, less human interaction. And it could swing the pendulum, right? It could get so much where customers are like, I'm done. I'm not going to deal with companies that deal with AI. Or you'll see some companies lean in a little bit. I think you'll see it both ways.

[00:28:13] But it's any place that you can infuse automation in workflows and connect systems and connect data. And at the same time, understand the data you have and insights to your data. I think that's the cheat code or the master unlock, I think is where you can see AI infuse itself. I mean, cloud code is a perfect example of that right now. It is a true unlock right now. Absolutely. Yeah.

[00:28:43] If you're not, if you're sleeping on that, you get on that now. If you're sleeping on it, yeah, I don't know what to tell you. But my guys are so excited. They are, I've got one guy that, he works a lot. I don't know if he's sleeping right now. And I know his wife's not happy. He's working around the clock, bugging everyone around there. He's like, it's like he's got cloud infused in his veins. He's just going around doing cool stuff.

[00:29:03] But, you know, you see, you see the spark in some people that were, you know, they're doing well, but all of a sudden they just kind of kicked it over because it truly does open up doors that have never been able to open this fast and this quick turnaround. Yeah. Yeah. So I want to get back to one thing you said around like this director of agents allowing our, at some point in the future, allowing your clients to create them. Yeah.

[00:29:29] I have an opinion why that matters or why we should be the enablers, because I think as an industry, like we did well with when all our job was to keep the blinking lights on. And then with the shift to cloud computing, others were faster, but we, as an industry, we adopted, right? I think we mostly missed out on SaaS, right? They just bought the CRM and didn't tell us. And so, you know, I think for some of us, this could be SaaS. We could miss out.

[00:29:59] And I know that they're definitely adopting things. So why is it important from your perspective that you guys are the ones to allow them to create an agent, choose from a directory, create an agent, that sort of thing? Yeah, I think it's important because we are their technology provider. We are their strategy, that we're their true partner, right? I think most, and so we service the SMBs, right? That's 50 to 500 employees. And we have, we go up as well, but that's kind of our sweet spot we focus on.

[00:30:29] And so a lot of them don't know where to start. And so you can see around the org, we've talked to customers that are using CodeWorker in the finance department and they're loading financial statements, right? Like, no, let's stop that now. And when we talk to them, like, oh, we're not using AI at all. And then someone's a finance, like, oh, yeah, we're using it over here. And they don't even know, right? So I think you get a lot of uneducated, you know, just don't know.

[00:30:56] They're not trying to do anything bad or, you know, malice, but they just don't know. And so you need a partner to help you with that. And by providing a safe, tested, guided system that will not take your data, that will keep it closed loop, that can operate in your environment, I think is important for a customer. But from an MSP, I think what you're talking about is we miss the stickiness.

[00:31:22] Like, CRM and SaaS, when it came out, it's not as sticky as it used to be because, you know, we've got the AI and whatever you want to call it nowadays, whatever, AI gate or whatever they're calling it, but a SaaS gate. But it's a sticky thing because I just call it two guys in a ponytail can't come down the street from another MSP and say, hey, you know, I've got this thing and it can work. And they're like, well, I've got this thing, but if I leave, I can't take this with me.

[00:31:49] And so it's so important to their business because it's super sticky. And I think that's where MSPs can really take advantage of the stickiness that we didn't get to take advantage of the SaaS because we're like, ah, it's just another thing. Ah, it's just like the internet. It'll be fine. But I think this time, the MSPs and the partners that will take advantage of it will win ultimately and will get the customers and will keep their customers longer than anyone else.

[00:32:16] Well, so I agree with you, but I also think that as an industry, we were great at infrastructure. Yes. We were great at buying tools and kind of being integrators, even though we didn't call ourselves that. You have a software development background. I do. Most MSPs don't. And so, you know, what you're saying, right, is you've got to do some things. Yes, you bought some tooling, right?

[00:32:44] It's not that you're unwilling to buy tools, but you invested heavily. And whether it's what you've done with automation or directory of agents or you allow them to create their own, it sounds like what you're saying is you can't just expect your vendor MSP to give you the tooling, right? To create that differentiated stickiness. You're going to have to do something different. I think so. Yeah. Listen, I'm a big fan of buying.

[00:33:12] You know, you may hear that I'm not because I have a software development background, but I like things that I don't, I can go to market quickly. But in this industry, in the size that we're at and the differentiation we're trying to create, buying is really hard because, you know, the mass tools out there, I won't name names, are built for, especially are built for smaller MSPs. They're not built for larger MSPs. So can you support our size? Can you support our agility?

[00:33:41] Can you support our demands? I don't necessarily want to be the largest client. I don't mind it, but can you support it? And that's the hardest part. People have a hard time with that because they're like, why don't we just buy it? I'm like, okay, we can buy it. We can get up in three to six months, but it gets you 80% of the way. And if it doesn't get you 80%, I don't know if it's worth it. And so we have to really think about it. But yeah, I think it's interesting.

[00:34:07] But I also think you use tools and you build on them, right? So you're using some of the main tools like a cloud code or using, you know, even just a Gemini or something like building on the tool or using the tool and integrating it into your system. So you're not really building things like from scratch. You're buying it and then building using the guard rails that you're allowed. But you're 100% right.

[00:34:32] The major tools in our industry today are struggling to provide a good, solid way and path to use automation and AI in the way that MSPs are. And where they are providing it, it's available to everyone. So it no longer difference. You can't, sure, you could use it differently. You could use it in different areas. You can make it more efficient. But there's not a lot of tools out there that do any things with customers. It's all internal right now. So that's a hard part too, right?

[00:35:00] How do you use something that's canned for our industry but use it for our customers? You have to actually build something. I'm only aware of one company that does AI as a service for customers for an MSP. And they're the only ones. And we'll see how that works out. But there's just not a lot out there, unfortunately. So you have to figure that out. And you've got to have the expertise. It's not easy. You can't just take an L1 and say, go learn this.

[00:35:27] Unless there are certain people that can take in technology really quickly. I've got a few of them. They're great. They come up in the range really, really fast. But there are so far in view between. And by the way, you've got to pay them, right? Because they're going to realize if I'm getting paid an L1 but I'm doing L35 work and AI and automation, I could go to another MSP or I can go somewhere else and make more money. So you've got to create that culture because this is not about money by itself. You've got to create a good culture and a working environment to get in the time and space to be successful.

[00:35:57] Yeah. Now, you point out a good, something I observe firsthand, right? If you get these amazing people that started out as an L1. And it used to be, you know, it took years and years to be an L3. And now they could, maybe they're not an L3, but they have a different career trajectory and they're creating this time savings and automations for you. And that's the thing is like I've even been caught by surprise of like, wow, we've got to, annual is too slow, right?

[00:36:24] We've got to really talk about how we put you in a different role, comp. Because otherwise you're getting L1 and you're creating like, you can do that L35, you know, value. And it's a good thing, but it's also like from a leadership perspective, I think it's a challenge because we're not used to, we're not used to that kind of pace. Not this fast, right? Like you said, L1, L3, it could take you a year or two, maybe three, depending on how you measure it. Is it certs? Is it how many tickets? Is it CSAT?

[00:36:53] Is it all together? I have people that probably couldn't do an L1 service desk work, but they can code apps in CloudCode. That's wild to me. Like you become a developer, you know, because Cloud is like a junior dev. It's not a senior dev by any means, but you could use it for junior dev. So you instantly become a junior dev overnight, which could take years in that aspect, depending on what you're building on and what you're using.

[00:37:22] And then with the help of a senior dev and some automation and QA, you can build a fully fledged app in a matter of a week or two. It's pretty crazy, right? But they couldn't take an L1 support call. Not anytime, any day. So you're 100% right. We're in an interesting time of the world and life and the management of it and the leadership of it. And I always say creating space because that's a big deal for me. And we have conversations constantly on my team.

[00:37:52] What do you want to do? My two questions. What do you want to do when you grow up? And what gets you out of bed in the morning? You know, what gets you excited? But those two questions mean so much when they tell you the real thing. I had a person tell me, I want to own a bar. Great. How do we get you doing that? Right? Well, for now, I just want to do this. Okay, great. But when you figure out that you unlock the root access to the people of what they want to do,

[00:38:16] it really is an interesting unlimiter of their abilities and what they can do for the business. And then you support them along the way. But this adds a new aspect and trajectory of a leadership that we haven't seen before. So we talked a few weeks ago, you had around 800. Now I think it's about 1,140. I think you have a really interesting view of this because of that.

[00:38:44] But also the other side is there's, even in small teams, there's a lot of fear where there'll be no L1s. Will my job go away? You know, and it doesn't matter if it's L1 or finance or sales marketing. Like it's, there's augmentation, I think, in all those areas. So there's fear in those. So I think you're like running a campaign, how to build better, right? Tell me, how do you address this and have you seen this? Yeah, even though we've addressed it and we've been addressing it for now 15, 18 months,

[00:39:14] we still have it a little bit, right? Because they're people, right? At the end of the day, we're all people. And so there's a fear deep down wherever that comes from at some point. But I think we've made a significant impact. So at the beginning of this year in February, we wanted to do kind of an AI week, but do like for three months long. So if you look up AI week, there's a bunch of companies. It's usually devs and they'll go for a week and use AI, come up with some cool apps,

[00:39:41] you know, they'll do a bunch of vibe coding and they'll just kind of learn and play and have fun. Because at the end of the day, developers are, you know, artsy people. They're creative people when it comes down to it. Yeah, it's a craft. A hundred percent. Yeah, people don't realize like it's just ones and zeros. I'm like, no, that's a little bit more than that. Yeah. So Build for Better is a great campaign where we offered it for everyone.

[00:40:02] And it's a three-month campaign of prizes, education, training, and an introduction of AI to our employees. Because we realized we didn't do a great job and we're trying to explain AI to our customers. And so we just need to do a better job of it. Well, if they feel comfortable, just like any other MSP, you're always can feel more comfortable talking to our customers about AI. And so we had a good amount.

[00:40:30] We had 40 people right now in the competition. So we have a competition, five categories, $1,000 each. The main category has like a trip to it as well. You could do a single person or as a team. And if you just submit an idea, though, you get 50 bucks. So we're just like, just submit an idea, you get $50. It doesn't have to be a good idea. No idea is bad. But we had ultimately 40 people say, I want to do something. So they're using AI Foundry, some cool things.

[00:40:55] Now, fast forward, if Cloudware was three, four months ago, we might have used Cloud Code instead. But here we are with AI Foundry. And so we've got some pretty cool things. And in the end, we'll have five. Everything goes in the backlog for us. So we'll put in the backlog. And I think we'll work on it if they can't finish it or make it better to where it's more enterprise-y for our organization or for our clients. But we'll have five products in the end that we'll productize and support and release to the wild. But it's more than that.

[00:41:25] We're giving training every week, two or three new trainings every week that's being released. We have prompt training. We have long training. We have short trainings. So just to get people introduced into AI more was important. And that was what we set out to do. So we're excited about the results. I think we're about three or four weeks left until the award winners. But some of the stuff is really, really cool. So I'm excited to see what that looks like when they're done. Yeah. Yeah, it's interesting, right?

[00:41:54] Because I could never predict who would be all in and who would create that spark and who would have fear and who would even know. There are a few people I'm like, oh, you're okay. I see you. I see that you decided to enter in. And you got a buddy because you didn't know how to do it. But so they got on a team. And so there are some people that definitely was surprising for sure. I want to talk about the right to go second or kind of client zero, as you call it. Oh, yeah. And I think that's important.

[00:42:23] I think you said something like first is fluff, right? And there's this vibe coding. And that's a whole other debate of what does that even mean, right? But even within non-vibe coding, we're rolling things out, right? We've learned, I think, that. So what does that mean? And why is that important, especially given the pace? Yeah, so the saying is we didn't want to be first. We wanted to be right. That was kind of the thing. So this is around client zero.

[00:42:52] So similar to other organizations like IBM or Microsoft, we build and validate new AI-driven workflows internally before we offer them to clients. So this essentially creates like a closed-loop environment with real scale, real-world noise, and real stakes producing more of a trustworthy and more supportable AI outcomes.

[00:43:17] And so it reduces our client risk and accelerates deployment velocity once the framework is proven. So if you look at how this operates, right? We kind of test it on ourselves. So put the oxygen mask on ourselves. You get on a plane, they always say, put the oxygen mask on yourself before you help others. That's what we're doing. We're putting the oxygen mask on ourselves first, getting all the bugs out, going through the pains, doing all these different things.

[00:43:43] And then when we're ready and it feels comfortable and it feels like we did what we needed to do, and it's where we want, then we'll go ahead and then release that to our customers, which we have. So we have two new AI products that we've released literally about a week ago.

[00:44:00] So like a secure AI chat and then a managed AI workspace, which is kind of like this secure AI platform that really enables workforce productivity, employee training, and AI governance. So it adds like an agentic AI piece to it and builds in the guide rails around that so you're in a safe environment. So those are things that we've kind of did internally and worked that out, and then now we're releasing those to our customers. Yeah, I think that's huge, right?

[00:44:29] I think if you rethink, not all of the experiments will pay off, even internally. That's correct, yes. That's even more of a reason why we want to be the lab. Right. Right. For our clients. Then just, you know, it shocks me once I've seen people go and not even create something, but sometimes buy something and then try it out on clients. You know, and you're like, you've got to have some expertise or some tracker here, don't you? Wouldn't you think?

[00:44:58] Like, so I think if you reframe what you're doing as almost like a lab, right, where you have experiments, then you need to run them. You've got to see, given your, like you said, scale, but also the reality of noise, because it's not always perfect. Right. Different users, different data, that sort of thing. You get to see a better idea if it works in the real world. And now you're framing this as like, there's products coming out the other end. Correct. Right. Right.

[00:45:25] That starts to sound a lot more like a software company or something that creates IP and just resales time. A hundred percent. Yeah. While we don't want to be a software company, we're very, very, very purposeful on that. It does, in my experience, being in software, I've been in SaaS, I've created products, I've owned software development teams, is very similar, but it's much more lightweight.

[00:45:50] You know, we're not building interfaces and we're not, we're building on products that we, that we have purchased to provide this type of premium experience for our clients. And it's just important for us that we're, we're doing in the way to, in the, in such a way that clients will appreciate this. We already have, we already have customers, so they've already purchased a week. We've already got a few customers, which is great. But, you know, we spent a lot of time. We spent a lot of time speaking to our clients last year, industry experts, prospects.

[00:46:19] We even hired a top tier consulting firm to help us better understand what AI means to SMBs. Because just because what we think it means doesn't mean that's what it means. And so we... Tell me what that means then. That'd be great if we can figure, probably figure it out. Yeah, well, we, we, what we did is we quickly found out they just don't know, right? At least during the initial conversation, like people were, SMBs, I mean, people in general, they're really struggling with AI is.

[00:46:44] Like they don't know how to, how to harness it or, or, or what to ask for help or what it can really do in their work environment that we talked about earlier on our internal. Like their, their concern is going to replace their role. And so we had to kind of take a step back and say, you know, how do we craft a strategy on how we can better serve our customers in the AI world? So, you know, we, we, of course, like most MSPs, at least I have my understanding that I've talked to them.

[00:47:10] We tried the simplest, simple solutions of deploying Microsoft Copilot in a SharePoint folder and enable the user to leverage defined information. And that's fine and good. And it works in some cases, but that's not really the true use case for AI. But the problem is, and we, we hit it on a little bit as data. SMBs don't have a great organized set of documents and folders. And so a lot of us can relate to that.

[00:47:35] And so the, the real value that we're, we are, we are hoping and realizing to achieve for customers really wasn't unlocked because the underlying data foundations were not ready yet. So, so again, didn't need to be first on this marathon of AI, but we wanted to be right. And we believe the services that we're offering now will really help empower our customers with the training, the tools and the governance they need to effectively integrate AI into their daily workflows in their business.

[00:48:06] What is your take? I feel like, you know, 10 or 15 years ago in the industry, we weren't, it was, you were, you were very uncommon to be sitting, you know, kind of at the board, right? We weren't really doing a lot of the VCO, VC. So we, we, we, you know, we, we were good, but we were, we were, our value was really IT.

[00:48:25] And then some of us have kind of, kind of started to act as a true partner and we all say it, some of us achieve it in some cases, but I think as an industry, we're still, I still think the best is we're, we're like a traditional CIO. And I mean that in a good way. It's like, you're, you're a good guy that can tell us if we can buy the thing or not, and you can keep the lights on like, but traditional in terms of like not really providing a strategy or, you know, guardrails or other things.

[00:48:54] So there's that, that it seems to be where we should be headed. I think my question to you is, you know, do you disagree on that? And the other part is, I feel like there's an unprecedented opportunity to actually make a meaningful impact to their business in a way that is much larger because like you said, there's cultural fears. There's all sorts of other things that are really not just IT.

[00:49:18] And if you can impact their sales or marketing or finance or whatever part of their business based on the tooling that you're offering, that's, you know, like, look, I have X budget to keep IT on and maybe I have a little bit more as we do refreshes or whatever. But it's like, wait, you can improve my EBITDA? Like I have a whole lot more budget for you now. Right.

[00:49:47] I think that's a strategy of security, but they still play a CIO role at the same time. But that works really well for us. We've seen that take off on our side and our customers really like those services, especially in the verticals of finance, financial verticals. That's a big one for us. We do have VCIOs as well for a little bit lighter touch, you know, maybe not as involved, maybe on the board, maybe not. But the VCSOs definitely. So I agree with you.

[00:50:12] I think the partnership and being ingrained in their business at the highest executive or board levels is a game changer for any MSP and I think makes it stickier, but also feels like a true partnership. Not just, you know, my IT guy over there I can call. And I think that VCSO and that VCIO is also, that role is changing a little bit as the VCSO is becoming a little more technical and they start to include the governance and the security around AI.

[00:50:42] Right. So being experts on security governance is great, but in the context when speaking about AI is important. So we're starting to see that shift a little bit also. So, yeah, I think it's important and I think to be a true partnership, which is kind of one of the things that I think we do really, really well as a partnership in that SMB. It's what we look for.

[00:51:04] We look for how can we be an impact to your technology or IT strategy, not just be your MSP or your IT people, but really be ingrained in your business, whether you're consuming VCSO or not. That's what we look for. That's kind of our pitch and what we do because we believe a partnership is best, not just, again, not just your IT guy you call. Hmm. So I want to get into a little bit of business model, right?

[00:51:29] Do you see like continuing to scale what you're doing now, especially the new AI as a service offering where you have a directory of agents, maybe the ability to create things, but basically you're kind of this not only IT company, but in some ways a platform. Or do you see a move to like outcome based, you know, we save this many hours, we're going to help you save this many hours or we're going to help you.

[00:51:56] I mean, ultimately, like the hardest one, but, you know, the easiest sell, right? You can tell me how much you're going to change EBITDA. Then I can certainly, you know, open up the ability to invest. Yeah, it's a great question. So we are definitely a platform. So we are moving from 800, literally as you're listening to this, we're moved to 1100 employees. We will be the largest international MSP that serves at SMBs.

[00:52:26] So we're moving international. And for us, keeping that platform while still serving the SMB market is super important to us. The concept of outcome based pricing is an interesting one. I'm only aware of today that I've talked to, especially in the last two months, all the conferences and executive boards and client advisory boards that I'm on. There's only one MSP that I know is doing it.

[00:52:55] And it's interesting how they're doing it in that they're basically just saying, do the things that I always wanted you to do. And if you do do that, I will give you a discount. So replace the old server, clean up your firewalls. Let us put our equipment in. And if you do all the things that we want you to do, because in MSP, we have this grand idea. Like if you do everything right, you will be amazing. Like you won't have any problems, right? Still do.

[00:53:23] But it is easier to support. It's cheaper to support, right? In theory. And so this particular MSP, that's how they do their outcome based pricing. Is basically, if you do all these things, you will achieve this NTR, this and this and this. And when you do, then we'll give you the outcome based pricing, which is really just a discount. And ultimately, that discount is just a discount they would have gave anyways. So it's what we're seeing.

[00:53:52] Now, someone may be listening to this and saying, I'm doing that. And if you are, I would love to hear about it because I'm super interested in this model. But it's nothing that no one's figured out the cheat code to give an outcome based pricing. Now, what I will say, though, with AI, this is an interesting space because I think you can open up the coffers, I think, as you kind of mentioned, or open up the EBITDA bucket of money by saying, listen,

[00:54:18] if we make this AI agent do what it's supposed to do, that will free up your two people in HR that you kind of don't need anymore. Or they'll come to that conclusion most likely. And if we do save you an FTE, we want X percent, right? Maybe 10 percent of that, right? Whatever. And so, one, it makes it even more stickier because now they're reliant on your solution for the replacement of their headcount.

[00:54:46] Or maybe they don't need a backfill or maybe they don't need a hire. So they're keeping their two HR and they don't need how to recruit. Whatever it is, right? But if you can prove out that value, I think there's something there. How you price that out, it's still TBD. I mean, AI is still new, especially for customers. For us, it's all MRR. We're not doing project services, right? I mean, we may do some projects, but for us, we're focused on MRR as an MSP as you should be. And so, if you use AI and use this, it's X dollars amount per user per month.

[00:55:14] And that's kind of how we're approaching it for now. But I'm tracking the outcome-based pricing really closely. Every time, it seems every time I go to a NABLE conference, they're always talking about it. And then I drill in and like, oh yeah, you're not doing really that yet. But it's an interesting concept and I think it could work, but it's got to be in the right framework, the right situation and how you prove that out is going to be hard because traditionally MSPs are not great at showing value.

[00:55:43] No, that's kind of what I was trying to, like whether, I don't know if we should do it. I don't know if it'll work. Like, it's interesting. I want to, if you're listening, I want to talk to you if you're doing it. Yeah, right, yeah. But I also see like the, like we say we're really providing strategic value, right? There's no better way to demonstrate that than avoiding hires or increasing sales or leads or margin or EBITDA or, you know,

[00:56:10] something that's truly measurable that, you know, hits the bottom line that anybody running a business can understand. And they won't, they don't need to care where you got that and which model powers it and all the things, right? Right. We won't care too much, right? Exactly. Yeah, it is an interesting model. I agree with you because you have to show value. Like it's hard to show services while you were sleeping, the things that you did, the things you do behind the background. I think it's a super important to have.

[00:56:38] We do that via our customer portal where we can show them hard data on their hardware, on their services, procurement, all the things that they do right in front of their face where it's really hard to show in other portals that are in the MSP market. So we try to show a lot of value that way. Custom reporting on their devices and their assets. So really cool things. But ultimately, it's still hard, right?

[00:57:03] It's not an easy thing and showing the right data to provide the right story of the value as well. I think that's something that's unique and hard to accomplish and achieve. Yeah. So with all this opportunity and depending on which headlines you read, AI can't do anything and it can do everything. Yeah. What's your take on like, is this a talent issue? Is it a prioritization issue? Like what's the constraint?

[00:57:30] Like where's your, I'd love to hear what you're thinking, right, Brian, in terms of how do I move the needle in the direction you want? Is it a talent constraint? Is it a prioritization, the number one issue? Is it something else? It's all of it. I think it depends on the organization. But at times we've had a little bit of, we've had a prioritization issue. We've had a talent issue. We've had a focus issue. We've had a dedicated resource issue, right?

[00:57:56] And so it just depends on the time of day and the time of season, if you will. It just depends on each one. I've heard them all. I've heard them all from other MSPs as well. Most of the MSPs I talk to, especially on the smaller size, it's a resource issue or a time priority. Because they only have enough people to service the customers they have in a certain way. And so peeling off a half of a resource, especially if they don't want to do it after hours, which can be hard if they're hourly. And so there's this complication there.

[00:58:26] And so I think it's really, really hard. My thing is start somewhere. I think, again, going back to what we started with is you have to start somewhere because I don't think you can realize the power until you do that. But I would start with automation and not AI. I just, I don't think, if you haven't done anything, I wouldn't start with AI at all. You also have this thing, which we haven't talked about, called tokens. And they're a hidden cost. I just saw Uber, $3.4 billion or million dollar.

[00:58:56] It doesn't matter. It was a lot of dollars budget for tokens, and they already went through it. And it's April of 2026. That, I think, is $3.4 million. Because they're just like, go code. Go use Cloud Code. And everyone went to town and ate up all their entire budget of tokens. And admittedly, they're like, I don't have a lot to show. People were playing around and creating stuff, right?

[00:59:21] So, with AI, it just brings in a different aspect and a different impact. You've got to show ROI. You've got to be intentional. So, if you haven't done that, don't worry about AI. Focus on automation. And then that could be a precursor to your AI journey. So, if you were starting over today, not maybe your size, right? But at the little bit smaller size. But what would you start? Like, you would start somewhere.

[00:59:47] And I mean that both maybe your favorite tool or two, but also more specifically, I think the mindset and the culture matter more than the tooling. And so, I'm just trying to figure out where would you start that might be a part of the business, that might be mindset or culture. I'm kind of curious. Where would you attack this if you were starting over? I'd probably think about this once a quarter. I am not the guy to go start an MSP, so I just want to be clear if any of my people are listening. But I think, like, what would that look like? What would I do?

[01:00:15] I can't tell tools because I might make some more partners mad. So, I can't name specifics. But I do have, you know, off-camera, off-recording, I can tell you. I have a specific PSA in mind or an ITSM. I have a specific RPA tool. So, I have specific things that I put them together. And I think, for me, it's all about data and it's all about interconnectivity.

[01:00:39] If everything connects together and is working in harmony and I can get at the data and everything is intentional in a way that is clean, which is very hard to do. But if you could do it right, then that's where I would start. I would start with a baseline and make sure everything is connected and I can get to the data. Because that's what we've done, but it's, you know, went backwards, right? We connected everything at the back end of all the systems.

[01:01:06] So, we created this data intelligence platform, which is kind of our hub of our systems. So, we've taken out, we've used ConnectWise. So, we've taken out ConnectWise as our center. Like, that's MSPs, that's their center tool. We've taken that out of the center and put it around on the outside, a hub and spoke model. It's traditional kind of data model, if you will. And so, our data intelligence platform is Microsoft Fabric. And so, right now, we've got 29 systems now. We integrated four more. We used to have 25 as of two days ago.

[01:01:34] And so, we have 29 systems now bringing in data into one data source, unstructured and structured. We marry that against our data classification and our data models. And we use Medallion models as well. So, gold, silver, platinum. So, I won't go through that. We could spend a whole conversation on that. But everything comes in and then I can convert them into semantic models.

[01:01:59] So, not only can I use my data now to show insights and to show cool things, which I'll talk about. I'll come back to that. But I can put them in semantic models so all my AI tools can use them immediately. So, all of our AI tools has instant access to our data internally to use to produce information. So, we have a RAG orchestration interface called Integers IQ. Think of it as like a chat GPT for us. But it's custom built.

[01:02:27] And we connect it to our, we call it the DIP, the data intelligence platform. So, we connect it to the DIP. And now, our employees have instant access to information that they've never been able to get or spend hours to get at. So, for example, they can be on a phone with the customer and the customer asks, Hey, my billing changed. Went up by five grand. Can you tell me why? Yeah, sure. They could literally type a prompt. I'm on with ABC company and their billing changed in March.

[01:02:57] Can you tell me why? Integers IQ will look at our system. So, it would look at financial, it would look at ConnectWise, it would look at all the different systems it needs to. And it would come back and does come back and tell them, Oh, well, change because of this project, which is related to a server. That server added this amount of project services. And then they also added these three users and literally on the phone, instead of saying, Well, I got to get back to you because I got to look at 12 different systems and come up with the reason why. They just found the answer in less than 15 seconds while on the phone with the customer.

[01:03:24] So, we're unlocking insights to our data that they've never really had in a quick, quick way. Sales reps are on the phone. Oh, yeah, we're competing against X, Y, and Z. Great. Tell me the last deals that we competed against and why we lost them. As we're talking to the customer on the phone. And so, sales gets instant access and instant rebuttals on how they can better sell that deal right on the phone with the prospect. So, we're doing these cool things where people don't have to run reports. They don't have to go into three different systems. They don't have to cobble together information as you do today.

[01:03:54] We have another unlock where you can unlock insights to the information. And now we've enabled, you can create PowerPoints. You can create Word docs all in our templates. So, if you want to load a bunch of words up and say, create me a PowerPoint, it uses our custom template for Integrus, and you can create a PowerPoint. If you want to load up customer documentation in the next release in, what are we, April? So, in May, we will strip out all the sensitive information. And you can load that up and we'll use all the LMs.

[01:04:24] You can choose which one. You can choose Claude. You can choose Gemini. Whatever you want to. It'll go out and get the information if you want to analyze it and come back in a safe environment. So, we're able to use these cool advanced tools because we have our data together. And that's why I go back. Long story short, sorry. I get excited about this stuff. Is that if you connect your systems and you connect your data and you make it in a clean way, and not even a clean way, just make it in a more organized fashion, you can really have an unlock to your organization that you've never been able to before.

[01:04:54] Mm-hmm. So, speaking of Unlock Sound, I kind of want to end this the way we started. Yeah. 35,000 hours last year? Yeah. Something like that. It's 30-something. It changes every day. So, I keep forgetting the number. But yeah, it's now well over. We're almost at 50,000. Almost 50,000. We're not even halfway through the year. I think you said the bottom line impact was... 8% EBITDA impact. That's huge, right? I mean, that's not...

[01:05:24] It takes a lot more of margin or gross margin or revenue or other things to make an 8% EBITDA. That's huge. My point is, look, we get excited about shiny things. I know I do. We implement things. They don't always have the payoffs, right? You're doing something that is really driving it. Obviously, anybody listening can tell you know your numbers. And so, where do you think...

[01:05:50] Like, how fast are you going to be able to spin this flywheel? Like, where does that end? Right? Is that all the juice? Is that really it? Or do you see this is, you know, the kind of improvements you're making that are real and dramatic? Like, where do you think that's going to go? Yeah, we definitely continue to automate things. There's plenty of things in the business we're automating. And we will continue to do that. As I said, we added an acquisition. Well, we just added it. We're going to add another one.

[01:06:20] So, we're going to essentially double in size to where we started by the end of the year. So, we've got a lot of things on our plate, right? And automating a lot of things and a lot of processes and technology is important. Our next stage, what we are really focusing on in 2026, of course, automations and continuing that, is around predictive values. So, really leaning into not just AI, but machine learning to come up with ways to better our business.

[01:06:48] So, addressing things like churn. We have a product coming out that we actually already released, but a final version called Integra Sentinel. Come up with these really cool names. But at least in my point, I'm not the marketing guy, but I think they're cool. So, Sentinel is basically exactly how it sounds. It's this Overwatch, essentially AI master agent that deploys many agents all over the business that looks at friction points of our customers. Or looks at, I was calling, like I mentioned earlier, the dark alleys.

[01:07:16] Things that we don't want to look down, but we need to because we're afraid what it's going to come up with. We want those agents to look down those dark alleys and come back to Sentinel and say, I found this. I found these friction points. We missed an alert. We missed an SLA. Things that account executives and our CX team don't find out until a quarter or months down the road. I want to know about now. So, you can get on the phone and say, hey, Bill, I missed it. We missed it. I want to let you know. Here's what happened. We're fixing it. We're on top of it.

[01:07:46] That relationship, that trust is just such a huge thing. I have a webinar on Thursday on some data that we, a couple surveys in the legal environment. And trust is like the number one thing, right? Between clients and a law firm. That doesn't go away with just a normal MSP either, right? Just a client itself. That trust is such a big thing. And that's how you build relationships. And like, oh, it may be reactive, but they knew about it, right? And they contacted me. They didn't try to hide it. They didn't try to gloss over it.

[01:08:15] And I knew about it. And I'm just waiting for you to call and tell me about it, right? And so, we're trying to do some things that provide valuable in a predictive way and bubble those up to the surface so that our employees can be more, provide an enhanced premium experience to our customers that they've never had before. But at the same time, get ahead of things so they don't fester. MSP, I mean, you know this, right?

[01:08:44] That's the worst feeling of like, why did we not know about this nine months ago, right? You don't want to be on a QBR meeting when you realize you dropped the ball. Yeah. And QBRs, those are other things. We're revamping those so that we don't have to do a QBR, but if we can put the QBR in our customer portal, that's just available. So, it's not a QBR. It's a daily BR, right? Daily business review. So, a DBR. And you can literally see the data. Now, when you have a conversation, we talk about strategy.

[01:09:12] I'm not having the conversation of, here's your SLAs, here's your MTTR, here's how many tickets. No one, they don't care about that. But now I can say, did you see the data? Oh, it was really good. Great. Let's talk about what you guys are doing next. What are you guys doing over the next six months? We talked about last quarter, are you changing that plan? And so, now we're, again, going back to your conversation, having those strategic conversations with customers are so important to a trust relationship. And I think if we can get a lot of those things out of the way and provide the value up front

[01:09:41] and the predictive value behind the scenes, we'll be an interesting MSP to force to be reckoned with. Yeah. Yeah. You'll be able to surface insights far faster. Yes. And what I love about that is so many people, there's fear about AI, there's uncertainty, there's fatigue, all of that, which I feel, depending on the day, right? But what I love about this is, you know, you're talking about basically building trust. Yes. Right? Not, not a scholarship relationship. Yeah.

[01:10:08] And that's something I learned, again, I didn't start my, my, I didn't start building computers out of the womb like most CIOs. I actually started in sales is what I wanted to be when I grew up for whatever reason. And, and so I learned early. You wanted to be in sales, like that's not allowed. Like, all right. It's just, no one at eight years old tells their parents they want to be in sales, right? My grandpa was a sales guy and he was, he was a very successful in the, in the, in the

[01:10:34] construction tools like hammers and automatic nail guns and stuff. And I would always go over their summers and he would teach me cool little things and little tricks. And so I just, it was, it was the, it was who I looked up to and I, well, I want to be a salesperson like my grandfather. And so I did, and I did that for my first five, 10 years of my career. But what, what was interesting, it helped me understand people, which is big, but it helped me understand the importance of trust.

[01:11:00] And if you trust, if you have the trust with your employees as a CIO and your executive team, yeah, you're going to miss some things. And it's not a quick thing to blame. It's like, oh, Brian's got it. We got, we have it taken care of. We'll figure it out. Right. And we can move on. And you have that relationship. You can always, you can always get past things when you have a relationship and trust. If you don't have that, you can't. And I think that needs to transfer to our customers as well. If they trust us and they know we have a relationship, they know we're not going to be perfect. We can't be, no one's going to, it doesn't matter how many things we automate.

[01:11:29] It doesn't matter how many things we predict. We can predict the future and it still doesn't matter. You're going to have a, you're going to fall on yourself, right? Any movie will tell you that. But if you have the trust and you have that relationship built, you know, you can go a really, really long way and do some really cool things together. Yeah, for sure. Well, I knew I could talk all day about this with you. And, but I want to say thank you so much for being on MSP Mindset, Brian. I love your candor.

[01:11:58] I love the way you're thinking, not just with data, but about people and improving the relationship. And I think that is huge. So thank you for being on MSP Mindset. Yeah. Thank you for having me. This was fun. I love, I love geeking out and talking about technology, but also the MSP industry. It's a, it's an interesting industry and the things that you guys are doing and highlighting in your new series too, I think will be really, really interesting to follow. But I appreciate you the invite and having me on. Absolutely.

[01:12:26] Are you open if people would like to tell you they're doing something or is there, what's the best way somebody could find you or connect with you? Yeah, I would say connect me on LinkedIn. I, I pretty much accept anyone. So all you vendors, I'll still accept your invites. Unless you look scammy, I won't accept it. So I would say connect me on that's easiest because if you email me, my email filters could put you in spam. But if you're more than welcome to email me, but LinkedIn me, message me, happy to talk, happy to help.

[01:12:52] You know, I can't tell you about everything, but I always, I always say, I can always tell you what I can tell you and I'll tell you what I can't tell you. So I'm happy to help any MSP or anyone out there with questions and guidance. And it's an interesting industry and it's cool. We get to help each other rather than, you know, most industries that don't want to talk about their IP or the things they're doing. And so getting comfortable with that, it's taken me a while, but I love doing, I love helping people and figuring things out. And I usually learn something along the way as well. Yeah, I love that.

[01:13:22] Don't miss out to take Brian up on that opportunity. You have a lot of data, but you're a wealth of knowledge. Thank you. Yeah, I love data, but I, you know, we do a lot of things, but I don't know everything either. So I'm interested if you're doing something cool, I'd love to hear about it. Yeah. Make sure to take Brian up on that. Yeah, for sure. Thanks, Brian. Thanks, David. Take care.