MSP Dumpster Fire Season 1 Episode 8 - Lessons, Insights, and Future Directions for MSPs

Welcome to the MSP Dumpster Fire series. This is season 1, the impact AI will have on your MSP. Welcome to the MSP Dumpster Fire series. This is season 1, the impact AI will have on your MSP. I am Robert Chiaffi, the CTO and co-founder of Progressive Computing, and I'm joined once again by my colleague and friend Alex Farling. Alex.

Hey, Robert, how we doing today? My name's Alex Farling, and I am a vendor in the MSP space over here at Empath. We're rocking and rolling and building out education for customers. And the wonderful thing about this series is that we've got to do a lot of educating. You know, you and I have gotten to unpack a lot of things. We've gotten to talk to MSPs about all the education that they need to do both for themselves and now for their customers. And in this episode 8, we kind of get to wrap it up and kind of put a bow on this thing and wrap this topic up for our MSP partners. So let's get into it. Yeah.

So, you know, just as a very quick high-level summary, you know, we talked about the history of AI. We talked about what AI is and what it's not, right? Trying to demystify some of the you know, the FUD and the fear around AI. We talked about how we educate and train AI, the various different models, right? We thought it was very important to arm everybody with a vocabulary of terms, at least just some basic understanding of these things. No one's expecting anybody here to go off and program these things or get in the deep innards of things, but at least having a fundamental understanding of them. We talked about the confusion, the downsides, and the blunders of AI, the mistakes that it can make, and how we need to make sure human oversight is an integral part of AI deployments. Then we got into AI jobs, specifically things around prompt engineering, which is an important— not only a job, but even just a basic skill, a bullet on somebody's resume. We talked about even things like using open and closed systems. And then the last episode, Alex, was around privacy, confidentiality, policies. We had that great interview with Mark to help us get into, you know, deeper into policies and making sure that we're protecting or putting up the right guardrails for our customers.

Yeah, and we're going to do another great interview today, right? I spent last week at the Kaseya Connect show and got to hang out with the guys from Hats AI. Jimmy and his team are doing really cool things over there. They've got a lot of time in this space. I think we would have to qualify them as significant experts at this point. So it'll be great to let— to step away and do an interview with Jimmy here in a little bit.

Yeah, Jimmy's super smart. Really like the guy.

Yeah, yeah. But, but before we do that, before we jump into that interview, let's just summarize and kind of talk about a few of the things that you're doing in your MSP, because I think that'll give some of our listeners some ideas, some things that maybe they can do in theirs, right? You've really used AI to kind of start a conversation with your customers. Can you tell us a little bit about that and maybe share kind of a high level of that video that you sent me last week?

Yeah, thank you for mentioning that, Alex. I went on my instincts, right? I just want to kind of remind everybody and also encourage everybody that may feel like, I don't really know what I'm doing with AI. And it's okay to say that, right? I didn't. And I still feel like I'm on that learning journey just like the rest of us are. But again, as a reminder, I've spoken about this before, how I decided to immerse myself in some self-education, reading, watching videos, just taking online courses as much as I could. And then I had a notebook full of notes. And said, what am I going to do with all of this? And then I started to get, you know, a couple inquiries from customers. And I'm sure everybody listening to this podcast or watching is probably getting the same. Hey, how do we use AI in our business? You know, my competitor's starting to use AI. You know, what are you doing for me as my IT professional? Those types of questions, right? So I just took a deep breath and thought to myself, what kind of value can I provide my customers? And I, you know, I came right back to the word education. What the path that I set upon, how do I then take everything that I learned, all this techno gobbledygook about AI technology, and how do I translate that into business terms and in valuable assets for my customer base, which are typically, you know, small businesses that don't have a technical bone in their body. And so that's, that's what I did is I built my own PowerPoint presentation. I organized all my notes and all my thoughts and everything that I learned and built this, you know, I did a voiceover in PowerPoint. And, you know, of course, I did my own little video editing. I'm not winning any Emmys or Oscars or whatever you win for your own little video. I'm not winning any of that stuff, but it's good enough. And I sent it to all my customers to say, hey, look, you know, here's a 40-minute video, albeit maybe a little too long, but so what. And I've had a bunch of people respond to it already, thanking me for it. Great education. Love, you know, what I learned, right? And I kind of prefaced it also with, you know, you don't, you know, there's not going to be a test here. Don't worry if I'm talking about something that felt a little too techie, right? I want everybody to sort of rinse and repeat on this topic. And so that's what I did, Alex, is I just felt like I needed to start the conversation. Simply by giving them some information. And it's been, I think, somewhat successful for us right now. I've got— I sent it to about 100 different unique individuals at various clients of ours. And I think we have close to 500 views on the video so far.

Well, you know, I think there's something important to be said here. And that is 25 years ago, I was a corporate trainer working for an outsourced an outsourced Dell tech support service provider. So you called Dell, 1-800-DELL-SUPPORT, and you got somebody who I trained who said, hey, thank you for calling Dell Technical Support. This is Bob. And we trained on Dimension and Latitude support. And it was what we did out of Dover, Delaware. It was a cool little business. But what they taught me as a corporate trainer was you don't have to be the expert. You just have to be 2 weeks ahead of your students. And, you know, we think we forget that educators sometimes are learning at the same pace that we are, just a couple of weeks ahead, because this stuff is moving fast. You won't be an expert in this. And I think it is worth it to say to the MSPs who are listening, you know, 2 months ago, 3 months ago, 4 months ago, Robert didn't feel like an expert, didn't feel comfortable having these conversations. So you put in the time and the energy, and now you're 2 weeks or 2 months ahead of your customers, and you're learning it a little faster than they are so that you can go in the room and have the big conversations. And it's still okay to say, you know, Mr. Kushner, but this stuff is all new and I don't have all the answers right now. Let me step away and get that answer for you. So I think we should preface that with, you know, preface this whole conversation with nobody's— nobody knows it all yet. And we may never, because this is a space that's going to be moving really fast for a really long time.

I love that mentality of bringing humility and transparency into that conversation. And it's certainly, you know, some of the things— and I'll talk a little bit in I know we're going to talk a little bit about some of the things that we're doing here as well moving forward. Uh, but if I'm on stage talking about AI, I'm making sure that that audience understands that I'm on that learning path just like they are.

Well, and I think the important thing is now, and we'll talk about this when we come back after the interview with Jimmy, um, there are folks in this industry you can partner with and everybody should be looking for the folks to partner with who spend their whole day in this space so that when you do get stumped, You go, but I got a guy and I can go have a conversation and I'll get back to you. And we can kind of queue up some of those, you know, hey, I might not know everything, but I'm networked and I know this is important to you and your business. So I've got a person who could help me with this part of AI. And I know you're going to give us some advice around that when we come back. But why don't we take this moment to step away? We'll do the interview with Jimmy and then we'll be back.

Great idea. All right.

I'd like to take a second to introduce my good friend Jimmy Hatzell to the MSP Dumpster Fire. Jimmy, thanks for making the time to do a quick quick little interview with us.

My pleasure. Thanks for having me. I'm excited.

Yeah, yeah, you've been around the MSP channel a long time, like as long as I can remember, between, uh, you know, some time at Scout, some time at CyberQP, and now you've got a new startup. Um, can you take a minute and just kind of tell the audience a little bit about what your startup is and kind of set that expectation or understanding of where you're coming at AI from?

Sure, yeah. So at HetzAI, we're in the business of putting MSPs in the AI business. So what we do is we find use cases for AI that work for small businesses, and then we productize those use cases and make it easy for MSPs to actually sell and manage and be the subject matter experts for it. So we are a software platform. We have a single platform where MSPs can build low-code, no-code AI applications for their clients. We have a whole suite of products coming out. We, we just launched the company this year, so we're new, we're in, uh, the early stages, uh, but we do have, um, a bunch of MSPs in our platform and active MSPs reselling AI managed services to their end customers.

AI managed services. Now I must be the old guy in the room because I didn't think those words would ever be a thing, but here we are.

You know, I, I didn't know it was going to be a thing either, and then I sort of uh, like early last year, I was like, I think this is going to be a thing. And then, you know, I convinced myself more and more enough to actually quit my job and start a company, uh, around that.

That, that is, that is the entrepreneurial spirit, right? We're just gonna put all our chips on the table and push them in, right? Um, so earlier in an earlier episode, uh, my good buddy Rob was talking to Mark Kirstein, and, um, Mark used a phrase that I thought was interesting. He said shadow AI. And I hadn't heard that before, but when he said it, I went, I know exactly what he means. He's talking about the fact that it doesn't matter what app I open these days, an AI prompt shows up somewhere. Whether it's I'm reading these questions off a Notion doc that I kind of built and, uh, you know, Notion's prompting me for AI, or whether it's Microsoft Word, or whether it's Google, or name your line of business application, there's some kind of AI in it somewhere. And if there's not, we can pretty much count on the fact that employees are bringing AI to the table today. It seems like one of the things you're trying to tackle is that shadow AI. Is that an accurate statement?

Yeah, absolutely. I mean, I like to tell this short story. I had my daughter's babysitter was over and my daughter was taking a nap and she's a college student and was doing some homework. And the whole time she's doing the homework, she had ChatGPT open and is going back and forth and back and forth. So I start asking her, I'm like, so what's it like? And she kind of explains to me that half the professors are like, use it. The other half are like, you can never touch it. And I was like, how many of the students are using it? She said everybody's using it. So, and then I think to like people who are in the active workforce, it definitely trends younger. More Gen Z workers are using unapproved AI applications because they see the power of it. I also talked to, you know, different people who do like SaaS monitoring, and very early on in the launch of these applications, they find high amounts of traffic to AI, you know, like ChatGPT, basically. And so people are using it, people are putting company information in it, and it's not secure, it's not managed, the outputs aren't controlled. So, you know, MSPs and their customers need to really think about What is the data strategy behind this? What is the policies behind this? So we, we, you know, we like to score our end users and MSPs on an AI maturity model, and the second level of the maturity model is people at our organization are using AI and it's probably not approved and there's no real formal policy on it.

Yeah, as, as, yeah, as, as we started to see the first conversations about AI and large language models, I went Well, there's going to be a business here, and it may just be our current content filtering companies that figure out how to block it. And this stuff is just like, it's going to be another problem MSPs have to deal with, kind of like, you know, malicious websites or, you know, shopping on eBay instead of doing your job or whatever. And then everybody went and plugged it into Microsoft Word and into every other app that you use. And it's like, I can't block Word. All of a sudden, life just got more complicated.

Well, The reason why it's so hard to block is because it's so useful and makes employees so much more productive and better at their job. So it's silly to just ban it and block it. You know, you need a strategy that's thought out and secure and managed and, you know, you have some oversight on, and that's what we're trying to do.

Yeah, I mean, it creates an interesting problem. I mean, Amazon just recently has been talking about this data leak from Amazon Q, right? Their product that is supposed to be their version of AI, and I haven't really kept up with it, but I was reading the article on Amazon Q and how it's leaking data for like locations of Amazon data centers and things like this that was fed to it during its training or during, during some of its, you know, some of the setting up of this model. Now Amazon's turned around and is actually issuing guidance to employees about how to use AI and how to use generative language models. And I feel like this is something that is going to come downstream really, really quickly. You know, quite often RPAs have been a thing in the enterprise space for a decade. They're finally making it, they're finally making their way to the, to the MSP space. I don't feel like we have a decade before we have to start having this conversation with our customers. What does that really look like for an MSP?

So what it looks like is, first of all, some training. You know, if you're, if you're listening to this, you're probably familiar with Empath. And I think it's extremely needed on AI, even just to have an educated conversation on AI with your customers. I think every MSP in the next 3 to 6 months should be able to have the AI conversation with their customers if you're not able to already, where you can just articulate the risks and sort of articulate what you recommend and what an offering is and what's safe to use. I think a lot of MSPs are moving in the direction of doing AI assessments. We've seen it already. It's sort of in its infancy. And there's, not, you know, really good standard way of doing it. We're trying to develop that right now and have some partners and expect some certified assessors soon. But think about how you as an MSP would do a security audit or security assessment to prepare someone for, you know, a NIST or aligned to, you know, NIST framework or something like that, or even prepare for a SOC 2 audit. I think the same types of assessments are going to be necessary and needed for AI. Because the quicker an MSP can articulate how a small business can safely and securely deliver 5 to 10 high-impact vertical and business-specific use cases to the end customer, the quickly they can move on and use AI in a secure way. Because without that, it's sort of— it's a free-for-all.

That assessment seems really complicated, though. It's one thing for me to go in and identify all the physical assets in an organization and make sure that they're patched. I can see that, right? I can do some scanning. I have tools available. I can't necessarily see as the owner what AI is being used in my organization, right? And you mentioned training, and obviously that's part of it. That seems like a really, really challenging assessment and one of those things where we're going to have to go in and talk to people at every level in an organization, not just go talk to the owner. I think back to the days of my Lifecycle Insights days where we're like, well, go talk to your customers about what apps they use and you'll learn very quickly. That the owner thinks they have 9 line of business applications, and if you go all the way through an organization of 100 people, you're going to find 100 business applications. Is AI any different? Like, and how do we find all the rogue AI that's existing out there in organizations?

Yeah, I mean, I think it's more about delivering the best use cases as quickly as possible than it is about identifying and cracking down on every AI out there. I mean, you do have to do the latter and I'm sure there's some automated SaaS monitoring tools and things like that that can certainly help with it. But employees are craving company-approved AI that they can use in their day-to-day. And employees who maybe aren't AI savvy or aren't, you know, Gen Z workers or nerds who like to play with all new tech, they want some training and they want some simple use cases too. So for example, If your company is a construction company and you're constantly trying to fill out RFPs and apply for state bids or local bids, zeroing in on that really high-impact AI use case and building out an AI that can respond to RFPs or create the first draft really, or at least, you know, create parts of it that's going to 10x your worker's ability or 50x or 100x to complete that specific task and do it better and faster. And it's not like this is going to take away my job. It's this is going to help me be better at my job. I think that is really an important step for the business. The way that we look at it on the maturity model is zero, you know, I know nothing about AI at all. One, people in my organization are using AI, you know, AI curious. No policies, anything like that. Two is, okay, horizontal use cases, things like generating content for social media, checking grammar, HR, KPIs, coaching, things that are sort of aren't specific to an organization. And in that stage is where you sort of introduce approved AI technologies to the end user of the business. And you can bring in, you know, something like Copilot, something like us, doesn't matter, both, right? But something that's managed by the organization. And then the next level is those vertical or industry-specific or company-specific, and we recommend getting there as fast as possible. The level after that is where everyone thinks that they need to go, but it takes longer than they, than they actually know, is when you have a company knowledge base that you're actually allowed to interact with an AI model. So say you have 500 files or 10,000 files or actual customer information that you want indexed and be searchable and, you know, AI applications to be able to look through them, that takes a lot longer and it requires proper access controls. It requires proper data validation and data scoring and, you know, figuring out what are the new versions of the file, getting rid of the old versions. And if you jump to that last step before moving up a maturity model, or doing the work that's involved, you're going to get googly gop, you're going to get data leakage, you're going to see a lot of the horror stories that a lot of MSPs have already seen, where it's like, hey, I turned it on. And then now a sales rep in my organization, it was searching for CEOs in the Miami area, and it pulled back the compensation package of the CEO of the company they work at.

Yeah.

So like, you got to be careful before you jump to that step. And that's where regular traditional— I'm saying regular, but security-focused MSPs, that's where their bread and butter is, because they sort of understand that stuff a lot better.

Yeah.

And where you can make a lot of money in professional services as well.

It's interesting, because one of the questions that I had on my list was to talk through kind of the privacy risk of AI and the risk of to the MSP of having all these undisclosed unidentified vendors that might have access now to data that they're not supposed to. I love your conversation about the maturity model because what you're talking about is slowly growing. And we have maturity models everywhere in the MSP space. This one just makes sense. I haven't heard anybody talk about it from this perspective yet, but it's wonderful to think about growing into a world where we actually can put sensitive data into an AI and not have to worry about it leaking because we're really in a world, right, where— and you and I both share Connor Swalm as somebody who we call a friend. He's in a really tough industry. In decades, training hasn't solved the phishing problem. And here we are with AI saying training is what you need to solve the problem. And so I always look at it and say, you know, if we give these folks a stick, they're going to poke their eye out. And so my gut reflex is to take away the stick. And the vendors have made it so we can't take away the AI stick. And so we really have to do kind of an all-of-the-above model here. And I think you're putting together something that if I'm putting it together correctly is a little bit of training, a little bit of knowing where you start, what your starting point is, and growing into an AI process or policy or maturity in your organization that suits your business and where you're at. Am I hearing all that correctly?

Yeah, absolutely. Absolutely. It's, you know, maybe instead of giving them like, yeah, everybody has access to sticks, but we start them with a twig and maybe put some goggles on and, you know, by the end there's a piñata hanging up that they can get all the candy from.

Right, right.

My old, my old— I guess he was kind of my, my coach at my MSP used to say, never give an end user a stick because— never give an end user a pointy stick because they'll poke their eye out. And if you give them a dull stick, they'll sharpen the end and then poke their eye out.

Right.

So I've carried that, that mantra with me for at least 20 years now. But as we kind of start walking through this right now is the time AI is here. It's in front of us. All the kids are using it. The old guys like me are going to figure it out eventually. It's time for MSPs to start having this conversation. What advice do you give to MSPs who are trying to figure out how to broach this topic with their customers? There's probably an outcome they're looking for. Some of us don't even know what that outcome is. What does, what does the conversation look like? And what's the thing they should be hoping to walk away from that conversation with? What does a good outcome look like?

I would say take your own organization through the process first. So, you know, what are your approved AI applications inside your MSP? What are some use cases that everyone can use? Hey, if you're, you know, I, I, there's, there's, I've met a lot of MSPs, nearly all of them struggle with marketing and aren't even, you know, aren't even doing any. So if you can just start using that, even if it's the act of generating content for your LinkedIn or your website in your own voice, and it's not googlygop that everybody else is putting out through ChatGPT, it's trained a little bit, um, It'll teach your organization how to use it. And then what are your AI policies and approved technologies inside your organization? And then move on. Okay, how are we doing ticket triage with AI? How are we using use cases that are specific to our organization? How do we take the level 4 engineer's notes from a ticket and translate them into simple communication that the customer can, you know, understand? And, you know, you don't have to do it the way that you'd handhold through an end user before you can get started talking to your customers about AI. But it's important to know this stuff by using it internally. As you know, probably, it's easier to talk about security when you're doing it yourself and you've used those applications yourself. Same thing with IT. It's a lot easier to have a conversation about how to set up organizational-wide Intune and SSO and all that when you've done it for your own company. So that's definitely what I recommend. And then I'd say, you know, start with find that customer who's willing to try new technologies, who's interested in AI and, you know, have a conversation with them and see what they want and what they're looking for and show them what you can actually do for them. And then also, like, I think like there's a big gap between people who know what AI is and use it every day to people who've never used it for. But then I'd also show an organization or a company or, you know, owner of whatever that you're close with that has no idea what it is, sort of what this technology can do, to get both reactions. Because in the first reaction, you're going to get, well, I want to go all the way to the end. Can it do this? Can it do that? Can it do this? Can it do that? And it's like, let's crawl, walk, run. Where if you show another company like, hey, look, it can generate a an email for you. They're like, wow, I, you know, my kid talked about this, but I never, I've never seen it happen. It types it on the screen like that. And you'll get some biases if you only talk to one type of customer. So that's why I usually recommend to talk to both and then, you know, work through an assessment with one or both of them if you can, if you're, if you're at that maturity level and see how you get on. And then, you know, reset and see how, what, what's a good way to roll it out to the rest of your customers.

Yeah, you mentioned the, the use case of an AI grant writer, and immediately my brain went to a nonprofit that I worked with at my MSP, and I'm thinking that would be life-changing for them. Like, that is someone's full-time job, just acquiring grants and making money to fund the programs that this nonprofit does. And I would be really tempted to run right to that end goal and go, this is the thing we have to do. But I think that is if I'm not wrong, if I'm hearing you right, that is the crawl, walk, run problem, right? We want that big fancy thing at the end and we've got to figure out how to get there without making ourselves a mess along the way and creating something where someone can go Google the CEOs or can go ask AI the CEO's compensation package, right?

Yeah. Well, it's, hey, this is what it's going to look like at the end. It's going to take us 3 months to customize this for your business. Here's the statement of work. First step is roll out AI internally. Second step is, you know, go through your current processes. Third step is to build out the first version of the grant writer and test it. And then fourth step is to turn it over.

I love that. I love that. And that is that— there it is, guys. Everybody says we as MSPs, we— all the advice we get is theoretical. That sounded pretty, pretty prescriptive to me. And I love that when we're able to just boil it down to something that simple. I think you've been kind of cheating ahead. It's almost like you're reading my notes. But one of the questions that I had was, Do MSPs have to worry about AI more so than their clients, right? As we, as we look at security, if an MSP gets breached, all their customers get breached. We have really sensitive data. We really are patient zero for a real outbreak. Is this just another case of where MSPs need to be client zero and treat themselves like a customer? I think you've kind of answered that, but do you have any more thoughts about that and how the MSP's use of AI may be the same or may be different from, from how a customer uses it?

It definitely is. It definitely is. And then the standardization is really important too, because if you have 100 customers on 13 different AI platforms here and there, it's going to be very, very, very difficult to manage and help them. So if you can, you know, go multi-tenancy standardization as much as possible, and then layer on as much security as, as you can, and then, you know, practice that same security and, you know, whatever, do, do eat your own dog food, I guess what I'm trying to say. You're going to have a lot, a lot better of an outcome. So certainly, I also will say, I think that there's a little bit of recency bias. And I don't want to downplay the security issue with AI because it is big and there is data leakage. But there's definitely recency bias where every MSP, when they started getting into security, it was so, so, so, so painful and ransomware was happening. Happening all the time, and business email compromise was happening all the time. In a lot of cases, it still is. Where AI— yes, there's a lot of new things. Yes, data leakage is a problem. But if you're using proper tools that aren't inputting your data into large language models for training purposes, it's not going to be as painful as that security journey early on was. I would say it's more similar to move to cloud in an accelerated pace than, you know, building out your MSSP practice on top of your MSP package and learning as you go.

Yeah. And, you know, all of these changes are— there's always the joke like this, the MSP industry has changed forever. It happens every 15 minutes. Our industry is just— the only true thing about our industry is it's always in some state of flight and some state of change. Um, this is just the most recent one, and by next year we'll probably have another one queued up. It just— it's just how things work, right? Um, but, but with that, I'd like to hear just your best guess. What's one big prediction you have for AI in 2025? Jimmy's crystal ball.

Uh, I think that we'll see the first AI-focused MSP conference. In 2025? I'm giving you more than one. I think that you're going to see a lot of the talking heads talking about managed AI and recommendations to roll it out. I think that there will be a handful of MSPs in 2024 who make a boatload of money rolling out AI-based projects or AI implementations into organizations and then in— and then increasing their margins, offering management fees and help with prompt engineering and things like that. But I think as we go into 2025, it's going to be sort of assumed that you're working on your AI package, your AI policies, and your AI strategy, both internally in your organization, but also externally to your customers. And what is the go-to-market that, that fits that and the pricing and packaging. So I think it's gonna be big in 2025.

Man, if only somebody was putting together a company that might help MSPs work their way through this challenge, because it sounds like there's a lot coming. I love, I love your statement that this is an opportunity for MSPs to deploy a new offering, to get more consultative with their customers, to actually go out and deploy project work because like it or not, MSPs are in a world of declining margins, a race to the bottom in help desk. It's a commodity product. It's, it's just not where the money's going to be. Um, cybersecurity is the thing that was everybody's last hot button. Um, it's, we're kind of on the waning end of that being your differentiator. Um, you know, everyone can claim that they're a cybersecurity expert now, or at least everyone will claim it whether they are or not. And so we're really in this place where MSPs are searching for the thing that's going to help that profit margin. And I think professional services is a big one. Professional services that lean on how you can help the end customer move money through the organization in a faster process. And AI certainly does that. These are the things where we win. So I think that's awesome. I love this conversation around AI. As we kind of get wrapped up here, are there any free resources you'd recommend for MSPs to go check out, you know, related to AI and kind of starting this journey? Because there's a ton of information out there and sorting through it sometimes can be the hardest part.

Yeah, certainly. I mean, we aggregated it, a lot of it together. If you go on our website and join our mailing list, we'll definitely send you out a lot of information. We do have some AI training available as well, which Hopefully I can talk to some folks at Empath and—

I know a guy.

Yeah, yeah. And get, you know, some or all of it into that platform.

By the time this releases, we'll probably be a whole lot further down on that conversation.

So there you go.

Yeah, yeah, yeah. We'll move that forward. Yeah. So what is that website? Where can people find that information?

hatz.ai. So just go on there, you know, sign up for our waitlist if you're— if it's if by the time this is released it's still a waitlist button. If, if it's— if there's no waitlist button, then there should just be a link to the training.

So there you go, there you go. All right, follow me on LinkedIn too.

I'm putting stuff out every day.

Jimmy, awesome. See, you've been one step ahead of me this entire journey. I was just going to say, if people don't know you, where can they find you?

Yeah, connect with me on LinkedIn. Um, I'm, I'm posting AI content, uh, every day specific to SMBs and MSPs.

We'll put your LinkedIn and your, uh, your website link in the show notes, make sure those are available for everybody. Jimmy, thank you so much for coming by and making time for us today. I know you're busy. Um, go address some of the things in that waitlist and we'll let you get out of here.

Great. Thank you, Alex. Appreciate you having me on.

Thanks, buddy. Man, Robert, what a great interview with Jimmy. That guy's really spent some time in this space and it's becoming obvious that, uh, he's going to be one of the MSP space's industry experts. Around everything AI.

Yeah, I think he already is. And, you know, Guy is— I just, I enjoy talking to him. And the more I do, the more I learn.

You know, I just love innovation in the MSP space. And having just been at a trade show and going to another one next week, I feel like you walk down the aisle and too many of the products and the vendor offerings look like just a cheap copycat of each other. And it's really cool to see somebody who's doing something different, something that is really original, something that is real innovation. And the fact that they're actually turning AI into a product that an MSP can offer to their customer, kind of make a sticky customer, create a relationship around something that is going to become integral to the customer's business, I think is super cool. I love innovators. I love innovation. So it's just really cool to see Jimmy and his team over there breaking the mold and doing something different.

I'm excited as well. Yeah.

But, but I think another important point is not every MSP is going to have AI products to deliver, right? At some point, being a generalist is the term that you used in the green room before we started this conversation. Like, as the IT generalist, when we've got to cover everything end to end for our customer, even if we don't have an AI product to sell, there's some other things that we really feel like you can do that set you up for successful conversations with your customers, set you up to kind of be the expert. Maybe there's even some partnerships that you can form. So we should kind of step in and have a conversation about that. And you had some good opinion. So why don't I let you kind of run with this? And, and we'll dig into what the MSPs should be thinking at this point, what they should be, who they should be connecting with, and maybe how to make sure that you have that deep bench of expertise at your, at your disposal. You know, I had a copier guy and a phone guy that I would go talk to at my MSP when I needed something that was outside of my purview. And, you know, I'll check the paper tray, but beyond that, let's go get the copier guy and he can do the fancy stuff.

Well, I love what you just said there, Alex, because it articulates very clearly for all of us exactly how we should be proceeding with AI. I think there's a bit of confusion about who has the accountability of AI, right? And I think as the generalists, as MSPs, I think our accountability is helping them steer, right, and strategize, you know, much like we may take on or help them, a client meaning them, help them select a line of business application. Are we software developers? I mean, you know, I know some of us are, but most of us are not. We don't write the software. We help find the right software. And like you said, we don't maybe install cabling, but we have a cabling guy. Some of us may not do phone systems, but we have a phone system guy. I just love that mentality. So let's embrace that with artificial intelligence as well, is that we may not be the direct purveyors of that technology, but we can help integrate that either into the strategy, into the conversation, and help them steer around it. Now, one thing that a little unexpected as I was going down my path locally here in my New York market is I was at a networking event of a couple hundred people. I've been a part of this group for 20-plus years, so I kind of know a lot of people. A lot of people know who I am as well. But I, you know, I ran into an attorney who I've been networking with for, for years, and we've just had a friendly kind of but casual conversation, no business exchanging. We got on this topic. And suddenly I realized that they're doing a lot of things around AI and even in cybersecurity. They have a professional team on staff that focuses on things like tabletop exercises, policies around AI. So, you know, long story short, got invited to lunch at their office, beautiful office, by the way, and they You know, we had about a 2-hour conversation. It was me and several of their experts about how we can partner together around this. And they likened me— it was kind of very interesting how outsiders look at MSPs. And they said, you know, you guys are like the plumbers. Like, we're going in and telling them, you know, what's wrong with their plumbing, but we don't know how to do that stuff. But you do, right? So I'm developing, and hopefully it will, you know, lead into something very fruitful. But the idea here is that we're going to partner together together where they're there. I can drive business to them around, hey, I have a set of customers that need policies, right? We talk about like acceptable use policy here, but I'm not an HR expert or a legal expert. I can kind of advise a client that they need a policy, but I can't write that for them. So that's where that partnership sort of comes in. And vice versa, they, with their set of customers, they can see, you know, the things that they need, but they, they've got enough knowledge to be able to identify that whether IT is being managed well or not, and then they can bring us in for the rest. So it's, I think it's a nice little partnership that I found with this specific company.

Yeah, and I think those can be really powerful, right? This is the opportunity to say, hey, I notice as we talk through what's changing in your business that you're dumping some data into AI models that may or may not pose risk to your business. You should at least have a company policy or maybe it's a terms of service with your customers. That explains how you use AI and what risk might be there, right? Right. How do we train people on it? How do we get people up to speed? There's a lot there. And the bigger your network— and I'm a huge networking guy, you know, I'm just big on, hey, I don't know all the answers, but I know somebody who's an expert in this little thing over here. And I think it's worth it for every MSP, even if you're the introverted computer guy who got into this so you didn't have to talk to people. There's an opportunity here for you to get out and meet some folks who are doing something similar, find some folks in the industry who are the experts, because it's going to be really helpful for you down the line to be able to say, hey, I'm not the expert on this AI thing. I know the basics, but I can bring somebody in who is specifically situated to help us with this particular part of the problem. And that's where relationships with guys like Jimmy, guys like Mark Kirkstein, who we interviewed in a previous episode.

Yep.

You know, just these are going to be super important to MSPs. And, you know, you kind of mentioned when we were in the green room again that walk in the vendor pavilion, right? There's, when you go to a trade show now, when you go to an industry event, you're going to see AI everywhere. Every single banner says something about AI on it. I walked by a 30-year-old anti-malware vendor, right? And their sign is all about—

No names, Alex. No names.

But their sign specifically says, AI native prevention for tomorrow's threats. And that just kind of breaks my brain a little bit because I know they have 30-year-old legacy code in this product, but we're talking— everybody has the AI buzzwords on their booth at a trade show. MSPs are going to be challenged a little bit. They're going to have to go cut through the noise and figure out what vendors are using AI properly, what vendors are actually adding value with AI versus what vendors are just, you know, sticking a little a little star in your application that when you try and type something that you want, want the application to do, AI pops up and gets in your way, right? And so we're going to need to understand the difference between the two. You know, I was literally typing something in Notion getting ready for this and the AI stuff's popping up, getting in my way. Did you want to ask me a question? No, I want you to get out of my way and let me do my work. But, you know, every vendor booth at the trade show has something about AI, something about, you know, machine learning, artificial intelligence, whatever. How can MSPs kind of cut through the noise?

That's an excellent question, Alex. So I'm going to answer this humbly from my experience as an MSP. And there's— I go through this and it's not a unique experience, but maybe, maybe isn't really felt by the vendors. So maybe this is a message even for some vendors out there. I walk in those doors, right? And my senses immediately become overwhelmed. With colorful booths and, you know, lots of people. And what often gets lost in the crowd, right, because everybody is vying for eyeballs and attention, are these little booths in the corner that still have that default, you know, uh, trade show, uh, banner that was meant just for them to find their booth so that that vendor can set up and you know, they, they, they barely have even, you know, laser-printed out of their office flyers that they can hand out. You know, they've got nothing really other than themselves. And, you know, just the table and chairs. What I'm saying to you is this, look for those smaller ones. And maybe I'm, you know, being a little dramatic there and saying, you know, how low budget they are. But the reason I'm saying that is because they're just starting out, right there. They haven't, you know, they're only 3 months old, but they may be an incredibly valuable vendor or resource for you to go talk to. So what I would encourage MSPs is the next time you're at a show, go look for those ones that are sort of stuck in the corner that haven't been around for 30 years, that don't have the big flashy booths or 6 representatives there at all times. Those may be sort of the next big thing. And, you know, I think it's our jobs to weed through some of them and find, you know, I think you can find some golden nuggets in there, right? So I've got like a, you know, Jimmy's further along, but I think of like Jimmy as an example of someone who just might be at a show by himself, quiet, not really, you know, you know, big booth or anything like that, but it's got something really awesome there to go take a look at. Right.

Yeah, no, I think that's super valuable. And, you know, the only thing I would add on to that is don't be afraid to ask questions. You know, I talked to a vendor the other day that told me that they're replacing SOC engineers with AI. And I think it's okay to challenge a vendor who says something like that and just say, hey, how accurate is your AI in replacing a human? Because what did I say? I sleep really well on the weekends knowing that there's a human that's covering my back. And when I asked this particular vendor that, they said, oh, well, we're getting there. I'm not ready to buy we're getting there. Not when it comes to my security, not when it comes to my SOC. So I think even as, even as the MSP, as the generalist, we have to know when to say, as the generalist, I still need a series of questions I can ask to say, how far along are you? Are we there yet?

Right.

And it's okay to push back and challenge because You guys know, MSPs, you understand where the risk to AI is. So don't be afraid to ask your vendors those questions. Just because they've got AI in their business name or in their product offering or in their marketing doesn't mean they're much more of an expert than we are at this point. In fact, it may just be a runaway marketing agent who's trying to, you know, trying to capture all the buzzwords that everybody's Googling for these days. Right.

And, you know, it's also okay not to be the early adopter. So I don't mean to, by the way, and I think you agree with me, Alex, Alex here is, you know, there are early adopters and you've got higher tolerance for risk and pain. And that's okay, right? If you want to dabble a little bit. But it's okay to kind of, you know, start a conversation with somebody like that, but not to engage right away and say, you know what, let's, let's talk again in 6 months when you've got some human oversight there or when you're—

Keep an eye on you. Yeah.

Yeah, yeah.

Or when your AI has become better trained with more data, right? And your data scientists have done greater labeling. See, if you've listened to our podcast, then you know what I meant right there when I said, you know, data scientists did more data labeling there. Yep.

Yep. I think that's a great point. And, you know, as we kind of wrap up this series, I just want to say it's been super fun. It's been really amazing. We've talked about a ton of different things and how, you know, relationships with the right vendors, like what Mark Kirkstein and Jimmy Hadsall are doing, can be super valuable, how just a high-level understanding of this can help you wrap your head around it and just be a little bit ahead of your customers and help them understand where this is going and maybe what the pitfalls are, that this isn't the secret sauce that solves all the problems that everybody ever had, right? I think it's really valuable to just understand we need to be paying attention to this a little closer maybe than some of us had. I know I started this as the AI Luddite and you definitely educated me along the way. It's been a. It's been an absolute blast. I think there's going to be some folks who watch this and say, Robert's my guy and I want to get to know him. Where can folks find you, Rob? And I'm going to make you spell your last name for you so they can find you because it's a tongue twister.

Well, we'll turn this into an Italian lesson so you know how to pronounce, properly pronounce my last name. Yeah. Well, thanks for that, Alex. Robert Cioffi, that's C-I-O-F-F-I. You can easily find me on LinkedIn, just linkedin.com/in/robertcioffi. Are chaffy. So there's no H in there. Although if you are in Italian—

I have done phonics— wants to put an H in there. I don't know what to tell you.

Yeah, no, don't, don't, don't, don't, don't. No H's in there. It's just like the word chow, C-I-A-O, that everybody knows. How can you say H? But, you know, all right, so quick little Italian lesson there. And maybe we'll do then series 2 of Dumpster Fire.

I'm gonna ask my AI why there's no H in there, um, and it's gonna curse back at you in Italian.

Or at least I'll make sure that that setting is enabled. Probably. But yeah, that's how to find me. I'm, you know, I'm kind of visible online, especially on LinkedIn. If you're at all involved in the IT Nation Evolve Circles, I'm a facilitator there. Or you could just look me up, Progressive Computing in Yonkers, New York. We're kind of easy to find. It's, you know, we're not that hard to spot if you happen to be local. Would even love to, a chance to meet up in person.

And there's a better than zero possibility that if you're at a trade show, Robert might be there. So, you know, good stuff.

There's a decent chance.

Yeah.

Yeah.

So, and so if you've been watching this, if you've made it this far, if you've made it to episode 8 in this series, we want to say thank you. It's been a heck of a journey for us. It's been an absolute blast. If you want to follow me and the team over at Empath, it's empathmsp.com or linkedin.com/in/alexjfarling, F-A-R-L-I-N-G. Ing. We appreciate all of you. Thanks for joining us, and we'll be back with another season when we figure out what we want to talk about. Thank you. This podcast was created by me, Alex Farling, and my good friend Robert Jaffe. This podcast is produced by Francis Belland. The music and artwork for this series were produced by artificial intelligence and edited by Francis Belland.