58. The AI Advantage: How Small Businesses Can Compete (and Win) Like the Big Guys with Brad Groux
Justin Shelley (00:00)
Welcome everybody to episode 58 of Unhacked. We are here today as usual with a special guest. And we're going to talk, surprise, about AI. We do that a lot lately. mean, like, guys, do you find anybody talking about anything else these days? This is a pop quiz, but.
Mario Zaki (00:15)
Hmm not really Now
Brad Groux (00:18)
Yeah, I think it's a...
Mario Zaki (00:19)
especially this time of year with like no sports except for like baseball. There is really no other conversation
Brad Groux (00:24)
Yeah.
Justin Shelley (00:27)
We're going back to sports. ⁓ Listen, Mario, we already shut that one down. Okay, so guys, we're gonna do some quick introductions and then we're gonna jump right in. I'm Justin Shelley, CEO of Phoenix IT Advisors, and I help people build their wealth through technology and then protect that money from who, Mario?
Mario Zaki (00:45)
Russian, Chinese, Korean, all the... ⁓
Justin Shelley (00:47)
The Russian hackers, the Russian hackers, ⁓
the government fines and penalties, and then the attorneys who will come and sue you if you do get breached. So we're going to try to keep all your money locked down as much as we are going to keep your ⁓ technology locked down. That's who I am. And I do business in Texas, Utah, and Nevada. Mario, tell everybody who you are, what you do, and who you do it for.
Mario Zaki (01:11)
Yeah, Mario Zaki, CEO of Mastic IT located in New Jersey. We work with small to medium sized businesses to keep them secure from not only Russian hackers, but the Chinese hackers, the Koreans, the, you know, every other country that Justin doesn't want to touch. ⁓ keep business owners ⁓ safe, secure, and give them the ability to sleep better at night knowing they're secure.
Justin Shelley (01:36)
Listen, think about it, Mario.
I'm trying to keep my attack vector as small as possible. I only want to fight one country at a time. You want to go to war with everybody. God bless you, but that sounds like a bold, that's a bold move. That's a bold move. So, today we are joined by Brad grew. Did I say your name right, Brad? Okay. And I know we talked about it before, but I suck at names and I already forgot. So Brad, thank you for being here today. Really appreciate it.
Mario Zaki (01:40)
Ha
World War 4 man
Brad Groux (01:54)
You did, yes, yes you did.
They have been.
Thanks for having me. Yes, so I'm Brad Gru, founder and CEO of Digital Meld. We're a B2B technology consulting company that specializes in AI and automation. We're based out of Houston, Texas, but we have people all over the US we work with. And our goal is, again, to allow you to use the tools that you currently already have within your business. So Microsoft 365, Google Workspace, those sorts of things. And punch above your weight class. Compete with the organizations bigger than you, because you can be more nimble, because you're smaller. You can wear multiple hats. You don't have to have analysis paralysis like a lot of the bigger organizations have. So that's what we like to do.
Justin Shelley (02:32)
I love that line punch above your weight class too. We're going to, we're going to dive into that. But first, Brad, I've got to throw a curve ball at you. A lot of people that I talked to from Houston and I'm, from Texas originally. used to live in Dallas. kind of, I'm all transient these days. Listen, that's what I'm getting at because every time I hear people talk about going to Houston, living in Houston, ⁓ escaping from Houston, there's not a lot of high opinions of Houston. How do you feel about your hometown Houston, Texas?
Mario Zaki (02:34)
Yeah.
Brad Groux (02:45)
Nobody's perfect.
I love it. ⁓ It's the most international city in the US. So there's, most diverse. So I love that. I'm, know, and it's the other than the weather during the summer, it's great. It's not a beautiful place. I will tell you that, you know, it's, it's urban sprawl. It's a lot like LA and that way, but we don't have the Pacific ocean or the mountains. So, but, and the cost of living is great and the food is amazing. So I'm, I'm a taco guy. So tacos and barbecue. So, you know, I, I used to travel for a living with Microsoft for five plus years and I've been all over the US.
Justin Shelley (03:04)
Okay.
Mario Zaki (03:17)
He
Justin Shelley (03:17)
Right.
Brad Groux (03:29)
spent many, hours and months and years on the East Coast, West Coast, and came back to you. Again, you can't raise a family in a lot of other areas, quite like you can here.
Justin Shelley (03:33)
and came back to Houston on purpose.
Okay, there you have it guys. All you I love tacos too, but all you Houston haters out there who keep telling me Houston's I and listen, I'm kind of ⁓ being playful here because I actually like Houston. I've only visited. I've never lived there, but I've never understood why so many people hate on Houston. So there we go. We're fighting back finally. Anyways, enough of that horse shit. Let's get in and talk about AI and cybersecurity. Brad with the, your experience. Do you have like a
Mario Zaki (03:41)
I love tacos by the way.
Brad Groux (03:43)
Yeah.
Justin Shelley (04:10)
Let's kind of catch everybody's attention. What is your your poster child for all things AI and automation specifically where you're talking about punching above your class?
Brad Groux (04:20)
Yeah, so, you know, I think most medium small businesses, ⁓ know, solopreneurs that, you know, time is your only finite resource, right? So whenever we go on, it's like, what problems are you trying to solve and what is the lowest barrier entry? So what's the low hanging fruit that we can knock out really quickly? The free for time to do more meaningful tasks. I think we can all agree that like meeting notes is like one of those things. And so you could use Fireflies, you could use Microsoft Copilot to immediately like summarize your meeting notes and be able to refer back and go to and chat with those meeting notes. Like, what did we talk about two weeks ago? Like, so you're just
You're running by your seat of your pants whenever you're just in a smaller organization. That's just the way it is. You don't have enough head count. starting with those, basically, let's capture your meeting notes, utilizing one of those tools. That's the easiest barrier of entry. That's how we kind of get people, give them a taste of what AI can do for you. And then we talk about things that are really common, really easy to implement. And so every business has an accounts payable process. They take in receipts, they take in invoices, they have time sheets, especially a lot of smaller businesses.
people that do contract work, they have customers and have contracts and so they submit time sheets and things. You have a lot of manual processes dealing with spreadsheets and time sheets and invoicing and stuff. And so let's start there. Let's start with that AP process. And again, the beauty is the AI has been trained so well by Microsoft, Google, Amazon on how to like read invoices and things. That's just the easiest thing you could do. And then can use like these automated tools like you have Microsoft Power Automator, make.com or other automation tool sets in the Google suite and such to.
that you probably already have, you know, and so just use those tools to start the automation. When you get a, when you get an invoice to your AP inbox, automate the intake of that invoice and the data entry of it. And then the only thing you really have to do is have human in the loop is what we call it. So you put eyes on it at the very end to say, okay, this is a legit invoice. Let's pay it.
Justin Shelley (06:04)
Mario, your thoughts on this and where are you using AI to revolutionize your life?
Mario Zaki (06:10)
Well, first of all, my thoughts on that is actually I'm like, holy shit, I need to do a little more of this because ⁓ I was going to actually ask Brad, could I integrate this with like my credit card and just have it, you know, pretty much give me, you know, a printout of what, you know, my biggest expenses are and my biggest ⁓ pain in the ass things that I need to probably get rid of, have it analyzed, you know, stuff like that. Could I have it just
pretty much pull stuff directly from my credit card.
Brad Groux (06:43)
Yeah, you can. So most of these large ⁓ organizations, have APIs that you can connect to with these automation tools really, really easily. And the big thing is, is most small businesses and medium sized businesses don't understand the power of their data, right? What makes Microsoft, Microsoft and Google, Google and Amazon, Amazon isn't their services they sell. It's all the data they capture at from those services. That's what's, that's their secret sauce. And you can do those same sort of things and get those same business insights, but you have to start capturing that data. have to lay that foundation and like,
One of those things is you have, you want to be able to see like how spending is across your organization. Let's say you're a small business and you have five, you know, 10 people total and five people have corporate credit or company credit cards. And you have to manually log into Amex every week to like just keep track of what they're doing. Now let's use the APIs available to pull that into like a QuickBooks and do some basic reporting. Um, and you can automate a lot of those, those tasks. And, we do the same sort of thing. Like this, like I was at an event the other day at a $40 parking for, and that's just the joys of downtown Houston parking.
for a Google AI event and like I just took a picture of it in my car, sent it to our APN box and the Power Automate took that and then helped process that into QuickBooks. And so it was literally a...
Justin Shelley (07:51)
So you have a lot of integrations.
So listen, when, when, ⁓ people get started with AI, there aren't a lot of integrations, right? You're, using chat GPT, or some version and you're asking the questions. I think most people have at least started there, ⁓ to do what you're talking about. You can't just open up a, ⁓ even a paid version of chat GPT and have it start talking to QuickBooks. Tell me a little bit about that process. How do we go from, let's call it from crawling to walking.
Brad Groux (08:21)
Yeah, so talk to a specialist like myself or others, but also use chat GPT. Say I have QuickBooks, tell it your ERP, tell it your CRM. You know, they have QuickBooks. I have QuickBooks, I have HubSpot, and I have five people with credit cards. How can I automate all of that together? Like how can I automate those sorts of things? And it'll give you, chat GPT is really good, and it reads all the documentation, the public documentation for all these services, and it knows their APIs, and so it can give you like a roadmap of what that would look like. Again, most of the time chat GPT is a
about 70 to 80 % accurate, I think we can all agree. And so you just, that's your starting point. And then you sit down with the people in your organization and, ⁓ and have those conversations. like to say, whenever we go into organization, I don't want to talk to the executive about what problem you're trying to solve. I want to start, talk to the people that are closest to the problem. So the people who live with it every day, because there's a translation layer sometimes that's happening on the golf course and that, you know, corporate events, and then actually the boots on the ground is what we call it. Like there's a difference there.
Mario Zaki (09:16)
You
know what Brad though? The one thing that I not necessarily disagree with but the one thing I want to add to what you just said. The problem is not too many people think that they don't even realize they have a problem because they don't think it's necessarily a problem. Like for me right now, I didn't really realize that I have a ⁓ problem but I know now that this is something that I should have done earlier and just have it automate a lot of stuff that I
Brad Groux (09:30)
Mm-hmm.
Mario Zaki (09:45)
have a bookkeeper do or have, you know, I'm doing manual, I'm downloading this, doing this, you know. So some people probably don't even realize that there is a problem that they can probably automate and make it, you know, something completely off their plate, you know. And as business owners, we understand the key to, you know, elevate and go to the next levels to, to, you know, automate as much as you can or have other people take stuff off your plate.
And AI is the perfect tool to do that.
Brad Groux (10:17)
Yeah, and I think it's a framing question. ⁓ what we like to do is we talk to folks, we say, what is something you do every single day that you wish you could automate? Like, you don't understand you have a problem. But if you're doing something multiple times, if there's a process to it, whether you have that process document or not, but if you follow the same steps, you can add and inject automation at least some point in that process. It may not be from A to Z, but it may be from A to and that's going to help save up and free more time.
Like, so start from those conversations, like what's something we do every day that we wish we could automate? And then literally see how can we start there? The funny thing is, is whenever I go to an organization, the first person I generally talk to, I think most people when you actually go into organizations, the receptionist, right? I asked the receptionist, like, what's something you wish you could do every day to automate? And I was in an engine, we had a company who's an engineering firm here and they have about 200 employees. And she was like, we have an executive conference room that has refreshments and food and all. It's a nicer conference room. said,
Whenever all the other conference rooms are booked, it's okay that someone can book that. But whenever there's an actual meeting for the executives or they're bringing in, you know, potential clients or customers, we want to make sure it's not done. And so I set up just an approval chain, basically, that people still have the ability to book that room, but it would go to an approval chain and she would be able to say, we predicting something? And then she would just say yes or no. And she could do it from Microsoft's team chat. She could do it from Outlook inbox. And it just, and I impressed her. And of course you impressed the receptionist.
Everybody in the organization is going hear about it because everybody talks to the receptionist.
Mario Zaki (11:41)
Yeah.
Justin Shelley (11:41)
Okay. This is a tangent, but like in the world of sales, which if we own a business here in the world of sales, I don't care what else you do. ⁓ I hear all the time, like, how do you get past the gatekeeper and Jesus Christ, don't get past the gatekeeper, make them your friend pro tip by Justin Sheller, me for more. ⁓ anyways, so one thing I want before we, ⁓ before I forget, we, we talk about punching above our weight class, right? And so as small businesses, we can now do things that
Brad Groux (11:47)
Mm-hmm.
Yep, I agree.
Mario Zaki (11:56)
you
Justin Shelley (12:10)
previously was only available to larger businesses. And I would argue you talked about paralysis by analysis. ⁓ I would call it paralysis by bureaucracy, but talk about what goes on in larger organizations where small businesses truly have an advantage.
Brad Groux (12:27)
Yeah, so I worked in enterprises my entire adult life for the first, you know, and I was a consultant for the last 12 years or so. worked five years at Microsoft as a PFE. So I would go to the largest organizations in the world, the Fortune 500 companies. was in over 100 Fortune 500 companies in my time at Microsoft. I was a platform's PFE. So I was an active directory clustering specialist basically. So I would go in and I would just scan their environment and see what's wrong and then tell them what they needed to do following our best practices.
And then I'd come back in six months and nothing would be done. And they had paid tens of thousands of dollars to get me on site. And it's generally because people are putting up roadblocks for the boots on the ground are actually trying to get work done. Right. ⁓ And so the problem is they don't understand the talent that they have. They don't free them up because most business leaders think in quarters. Because why? Because that's where their bonuses are based around where the people on the ground, they're thinking like, hey, I'm to be here five or 10 years. I want to make this environment as best as I can.
And so we have to change that mentality of like you're laying the foundation. Now you won't immediately see ROI sometimes, but you have to, you have to look into the future and say, and predict like, okay, we don't, we're not going to adopt AI wholly today. It's not going to be all across our own organization, but in five to 10 years, it will be. And so what do we need to do doubt now to lay that foundation? Cause if you don't set that foundation now and the foundation I'm talking about is data. Your data is in disparate systems. Your data is in people's inboxes. It's on people's desktops. We still know that still happens, right? It's in people's heads.
⁓ and the baby boomer boomer generation is about to retire. A lot of them are, and we're going to lose decades and decades and decades of business knowledge. We should start capturing that now utilizing tools like Microsoft, Copa and other automation tools is like, just start having conversations with those people and capture that information. and again, you can punch above your weight class because all the same tools that these large organizations have, you have access to. There's it's democratized. You know, ⁓ I always say technology is a great equalizer and it really is. ⁓ we work with construction engineering firms right now.
And Jacobs is like the engineering for they're an $18 billion construction engineering firm and they use the Microsoft platform. They have a platform called alluvial that uses power BI and automate and like Microsoft fabric and stuff. A small business with five employees has access to that, that same platform. And they, they just have their processes that they've developed over decades, but there's no reason you can't get up to speed too.
Justin Shelley (14:45)
Alright so I've got a
Mario Zaki (14:45)
And you
go into those type of businesses, five users, kind of get them up to speed. Now, something like that, how long does it usually take?
Brad Groux (14:56)
It all depends on that, you know, again, because small businesses, they're also trying to run their business at the same time too. So they don't have time to sit for eight hours in a conference room for two weeks. Right. So it's, we work with what we meet you wherever you are is like, Hey, I have two extra days this week and three extra days next month. Let's just start that process. We'll get that roadmap. We'll have a quick meeting. We'll understand like what our goals are. We'll define those goals and we'll work with you. However you need to be worked with. Um, and the goal again, we'll start with those smaller.
wins so we can build that momentum. You can see the results of the work you put in. And our goal is for these smaller businesses to be able to hire more people because you actually have more time to go sell. You're not doing your processing your invoices. You're having more time to grow. ⁓ We also want to make a point that we use automation not to lay people off. We use automation to enhance your business and make you grow. ⁓ You're trying to increase productivity without necessarily having to increase headcount.
Justin Shelley (15:51)
Listen, I want to, I want to talk about that for a second. ⁓ it, it, to truly use this, I believe you have to already know your, your industry. It's not like I can come in and be a brain surgeon through AI. ⁓ so the other day I'm working on my, my PSA and this is what it companies use to, you know, for ticketing, for communicating with their clients, for invoicing and all that. And I needed some insight. needed some customized reports that doesn't come with the system.
And it took me a while to realize this, but I could ask AI, just like you said, hey, here's what I need to do. How do I do it? This is the system I use. ⁓ Get me started. And it walks me through. Now, first of all, I will tell you it's working. said that they, know, AI knows the documentation and the API and all this stuff. Mine doesn't, mine's stupid, I guess. Actually what's happening is it's working off of an old set of documentation. The user interface is newer than AI has access to.
Um, and I haven't solved that problem yet, but I'm getting there anyways. So we're, we're custom. Let me finish that. Come back to that. Cause I do want to hear your take on it. Um, I'm, I'm writing reports, which really means I'm building SQL queries. Okay. Now I have a background in this. used to code for fun. I love it. It's a passion of mine, but it's been a decade or more since I've really done it. I don't know what the hell I'm doing, but I can tell AI, Hey, this is what I want. Give me a SQL query. go into my.
Brad Groux (17:00)
Mm-hmm.
Justin Shelley (17:15)
my PSA system and I punch it in and it breaks. But here's the kicker. I can figure out why it breaks or I can plug it back into my AI and I can understand what's going on so that I can translate between these two systems because I don't yet have them tied together perfectly. ⁓ But the point is, it's a long story to say, if you don't know what you're doing, AI can kind of help. But if you do know what you're doing and then you add AI to it, it can turn you into a goddamn genius, right?
That's my experience where I'm doing stuff that was not available to me. Not very long ago. I have a limited bit of information. I understand the system, the concepts, but I can't get past that until I bring in AI. So that has been kind of a game changer. Now, ⁓ react, but tell me you, you, you tried to cut me off. Go ahead and tell me what you wanted to say there.
Brad Groux (18:03)
So whenever you find that your GPT doesn't realize what you're talking about or it's looking at an older set of data, what I literally do is I go to wherever that new data is, I print as PDF and I upload that to my chat.
Justin Shelley (18:14)
See,
and if my PSA had good documentation, I would absolutely do that, but they don't. And I got in a fight. No, I got in a fight with ChatGBT. I'm like, listen, I'm using version this. You're using version that. Can you fix it? And it says, oh, of course, I'll update everything. And still gives me the same old shit that I was anyways.
Brad Groux (18:18)
Okay, yeah, yeah, the...
Yeah, so people don't realize it's called generative AI for a reason. It never tells you no. It'll make up BS. That last 20 % is probably BS. So like always double check, have that human in the loop. ⁓ Yeah, so it's great from the perspective of, know, it really is. In the future, there's going to be people and there's going to be people plus AI. And the people who, regardless of what you do right now, I talked to someone
Justin Shelley (18:38)
Yeah. All right. Yep.
Yeah.
Brad Groux (19:00)
on my podcast this week, he's in Hollywood and he's a producer and he says, you're going to have to adapt to use these tools. Like you can pick it and complain and all those things like that. Because if you don't do it, our competitors are going to do it. And if our competitors aren't in the United States, it's going to be the Chinas of the world, right? And so it's going, you're going to have to adapt it. This is happening whether you want it to or not. So I encourage you, regardless of what you do, figure it, go to chat GPT and say, I have four extra hours a week.
over the next six months, I wanna train up to learn how to integrate LLMs into my workflow. And here's my job title, here's my resume, here's my LinkedIn page, what can I do? And it'll give you that basic roadmap to kind of start there. ⁓ And then use these tools to help elevate yourself. ⁓ We were talking about these things, so I don't remember the actual question, but hopefully that gives you some guidance.
Justin Shelley (19:49)
Yeah, no, that's that's perfect. So I want to let's let's kind of wrap up and then pivot. So Mario, any final thoughts on just using AI in the world of business punching above your class doing great things that you couldn't do before? ⁓ And then we're going to move on to security around AI. So Mario, any final thoughts here?
Mario Zaki (20:08)
I mean, for me, really is ⁓ a like we were in the business and we work with this stuff all the time. You know, I use it also with our PSA. You know, we, we use the same PSA. actually has a built in AI feature that can go and check, you know, previous tickets, check previous knowledge bases that you've uploaded and tell you like based on this ticket, this is what you should be doing. And
Sometimes the one thing that I find myself, you know, 40 something years, 45 years of, you know, life, I completely forget sometimes that, you know what, why am I doing this manually? I just throw it in chat GPT and I'm like, you know, and, and I love, I love having conversations with it, you know, like, because we sit there and we go back and forth. My wife hates it. My wife is like, you'll talk to, you'll talk to chat GPT, but you, won't talk to me. And I'm like,
Justin Shelley (21:04)
You
Mario Zaki (21:04)
why
don't we just cut the middle man? Why don't you turn on your chat GBT and talk to it and then I have not have to bother me and she refuses on doing it. You know, but you know, other than that, it really is something that is easily available and the power of it is amazing. It really is.
Justin Shelley (21:21)
Great relationship advice by Mario Zocchi. Follow us for more tips.
Brad Groux (21:24)
Yeah, yeah. And one thing I also mentioned is so you brought up that it's hallucinating, know, with because basically how for those who don't know, LLMs work off of context and the context that they've chat to be trained on is literally they scraped the entire Internet and we can talk about the ethics of that later. But that's how they know how it's how it knows what it knows. So if you want to limit that context window, use a tool like notebook LLM from Google notebook LLM dot Google dot com.
Justin Shelley (21:29)
Mm-hmm.
Brad Groux (21:47)
And it's only going to give you context on whatever data you upload. It's going to give you responses and so I do this from like white papers. It's like understanding. I can't keep pace of all the things going on in the AI space, like from the research perspective. And so I have a notebook LM like workspace where I go upload these 90, 120 page, 120 page, you know, research papers. And so I can chat just about those research papers and learn more about those research papers and that information. Do that from the perspective of, I want to hone in on why I'm having this issue with this little specific product.
And it's not going to go out to the internet. It's only going to be based on what you've trained it on with the uploads you have. And so there are options to like fine tune there too. ⁓ But you need to know that, hey, yeah, you are going to get some BS answers sometimes. And that's why it's really important for you to understand your business. And I tell folks this all the time is like, I again, work with construction engineering companies all the time. And even if they're in the same size and the same market, how they operate internally is completely differently. So I can give them a framework of like what their processes are probably like.
Mario Zaki (22:22)
You that.
Brad Groux (22:46)
you know, they're from like 70 to 80 % accuracy, but the last 20 or 30 % is their secret sauce and how they work individually and internally within their organization. And AI is really no different.
Mario Zaki (22:56)
Now, Brad, I do have a question for you. ⁓ Because obviously people are scared of this and stuff like that. you mentioned a few minutes ago, you're not trying to get people fired or lose their jobs. You're trying to get them to work smarter. ⁓ Based on the clientele that you have worked with, how much of them have actually been able to automate and actually let people go?
Brad Groux (23:22)
So we've had a couple of people let folks go like at the lower level at the basic, like the, let's say the AP automation process, the people that literally did not, they were just doing basic data entry and they're like part-time workers. But our goal whenever we go in is like, hopefully we want to also, if they lay people off, really bad at also telling about how they can easily use AI to upskill those people. The hardest thing is business owners. think you would both agree. I would love to get your insights is hiring good people. That's literally the hardest thing we do. Maybe other than like sales, like closing a sale. That's really hard too, but.
Hiring a good person, someone that you can trust with your company, especially when you're a small business, I can trust this person. I know this person's gonna do the right for my company and be an ambassador for the name of my company. That's the hardest thing you can do. And that's the beauty too is like in the enterprise space, it's really easy for the elder is gonna lay off this whole division. And mid market and smaller enterprises, there's more of a family feel. hopefully we have those conversations, but ⁓ folks in like the manufacturing industry, folks in the ⁓ transportation industry,
⁓ I would encourage you now to start upscaling. And again, you don't need to change industries, but say, how can I utilize these tools to make, you need to make yourself indispensable. And the only way you can do that is your human plus AI.
Justin Shelley (24:32)
So I got two points. go ahead, Mario. Okay. So I know who I am, even though you don't. ⁓ So two points. What the, the case I was mentioning before where I'm using AI to customize some SQL queries and whatever, I will tell you that because of my background, my knowledge, my skillset, I could do it. There is no chance that any of my clients could have done what I was doing, even using AI. Like it was way, way too complicated.
Mario Zaki (24:33)
Can I? No, go ahead Brian, go ahead. Go ahead, just answer.
Justin Shelley (25:00)
to the point where it was hard for me to do. It took four hours. What really, if the AI was properly trained, honestly should have taken minutes. ⁓ But so that number one, I think to really use it in a way that really does indispensable, fine, more productive, more valuable, however you want to frame that, you have to start with some basic skills to be able to even know what questions to ask or even where to start. So maybe we get past that at some point, but we're not there today. Point number two, and this was a fear before AI came onto the scene,
I don't know. It wasn't a big movement, but you hear people running around, their hair on fire, screaming the world's ending because for the first time in human history, our population is no longer increasing exponentially. And I don't remember the exact stats. can't tell you what the numbers are. I don't know if we peaked and we're actually in a decline, but the fear was, ⁓ the, population with its lack of growth, we are going to be as business owners, we are not going to be able to find enough employees.
Our biggest problem is going to be that we don't have the people to do the work that we need done. Now everybody's freaking out like, my God, we can do more with AI. So everybody's going to be out of work. Like, is it? It, it, it, and maybe, maybe these two and my crystal ball says these two balance each other out because we are getting to a place where we're not, the population is not growing the way it once was. ⁓ and that's scary to economists. Like if, if, ⁓ if a country's economy, the GDP or whatever, I'm not an economist.
Mario Zaki (26:08)
you
Justin Shelley (26:28)
Full disclosure, but if we aren't growing that causes major problems I mean especially at the rate that we take on debt, right? So this this is one problem that I haven't yet heard anybody talk about that AI Probably is gonna solve much more likely than AI is gonna take all of our jobs We're gonna be sitting around with nothing to do thoughts on that
Brad Groux (26:35)
Mm-hmm.
Yeah, so we've had an engineer shortage for 30 years. Like in the United States, we've had like mechanical engineers, electrical engineers, we've had an engineer shortage for 30 plus years. And I being from Houston, Texas, and I grew up in a city called Deer Park where all the it's Carson and Genet Coast, it's where all the chemical plants are. And so I worked around like really hard working six figure making blue collar workers. So I think there's going to be a rise in skilled labor. think, you know, there's there's a skilled labor is something for apprenticeships.
Justin Shelley (27:06)
Mm.
Yes.
Brad Groux (27:17)
You know, in Houston, you can graduate high school, go be an apprentice and make 90 to $100,000 a year as an apprentice. And then by the time you're 21, you're making $150,000 a year. You're working in a chemical plant, there's risks involved with that, yes. But then you can, they have amazing benefits and things. And so I think again, the younger generation need to know that it's sweat equities, okay, still. think we're, gone are the days of getting a college degree for four years, then working in the same company for 35 and 40 years and retiring.
Justin Shelley (27:43)
Sure,
yeah, that's gone.
Brad Groux (27:43)
We're in a life,
we're in a lifelong learning scenario, folks, like you have to continue to evolve and grow. And I agree there are countries like Japan who have massive like birth rate problems. A lot of countries in Europe do as well. ⁓ I think it like you said, I think there's to be equal out. Nobody really knows there's no crystal ball. We don't know how it's going to Yeah, yeah, no one really knows how it's going to affect. But I think we can predict the trends. The trends are if you are a lower skilled white collar job,
Justin Shelley (28:02)
No, there's not. Well, I mean, I have one, but I just keep it pretty quiet.
Brad Groux (28:13)
your job is, you know, if your data entry, something like that, your job is ripe for being taken away and being automated way, especially work at larger companies where there's no, you know, no, no real ethics decisions involved. They're just looking at you as a one and a, you know, one or zero. So just know that you should probably start up skilling now and use chat. You can say, this is my goal. This is what I do over the next six, 12 months. I only have two hours a week. How can I do that? Like time is the constraint, right? So ⁓
Justin Shelley (28:36)
mean, you don't, yeah,
you could, you could literally just say, here's my situation. What should I do next? You know, you don't even have to give it your, your proposed answer. Let it throw out some ideas at you.
Mario Zaki (28:47)
Yeah, but you have to program it in a way where it's not gonna just keep agreeing with you. know, like tell it, like find me the problems, you know? Because no matter what you throw at it, they're like, yeah, this is a great idea. You you should jump off that bridge, you know? ⁓ It's definitely... ⁓
Justin Shelley (29:04)
I, you know,
I, I'll be honest, Mario, I hear that, but, ⁓ and maybe it's the way I programmed mine. Maybe it's just the way I talk to it. Cause it does. It's a mirror, right? It just looks at us and I'm a, I'm a no bullshit kind of guy. So, ⁓ my AI does talk back to me that way. If I say, here's what I'm thinking, it will come back and say, that's not a great idea and why, and here's a better option. ⁓ so yeah, do be careful if you're, if you're AI, ⁓ companions slash invisible friend. I don't know what these things are these days.
Mario Zaki (29:13)
Mm.
Justin Shelley (29:31)
I do know that if mine was taken away from me, I would be very, sad and very lost. But anyways, ⁓ if yours is always like telling you how amazing you are and you look around and you're really not that amazing, ⁓ maybe talk to your GPT a little bit different.
Brad Groux (29:43)
Yeah. And use the GPTs to learn how best to use those GPTs. So not all models are created equal as well. Let's say you're just, you've got an itch, you've got an idea and you don't know about the market research. You don't know about all the different parts of that. You know, you have a new business segment you want to reach out to. So go to chat GPT and talk to 4.0 that's for the letter O and say, Hey, I have this great idea. I need to write a prompt for deep research. And it'll give you a deep research prompt for that. And then you change the model to the deep research model, which will go work for 10 or 15 minutes.
Justin Shelley (29:47)
Correct.
Yes,
yes.
Brad Groux (30:12)
And then it'll provide
you a 20 or 30 page response. And then once you have that deep research response on ChattyPetit, go to Google's Gemini and they'll give you deep research for free as well. So go to gemini.google.com and do their deep research and compare and contrast. then ⁓ don't, yeah. Yeah, and so again, we're all trying multiple things every day. We're all.
Justin Shelley (30:26)
And use AI to compare and contrast too, because it'll build you a table of differences.
Mario Zaki (30:28)
Yes.
Brad Groux (30:35)
I think in the future, you know, I had a conversation earlier. gave us talk earlier this month, but basically we still teach kids like we did in the 1860s, reading, writing, arithmetic. The we've upgraded the curriculum a little bit, but we still have that same process. I think in the future we should be teaching people how to be problem solvers, how to have soft skills. You're going to, all these tools are going to take away the need to know that the basics of that. So it's like having faith in your ability to figure things out is the single most important thing you can do. And that's whenever I'm a hiring manager, I ask people like,
What problems have you solved? I give them a problem set. I like, how would you go about solving that? And even if they can't solve it, I want to see how they would go about solving that. That's the most important thing to me. You can learn technology. You can learn processes. that's faith in your own ability to figure things out you can't learn. That's earned.
Justin Shelley (31:23)
⁓ Mario, any thoughts or questions?
Mario Zaki (31:27)
No, I was thinking, mean, you know, with the decline or not as growth as your, you know, rate that we were mentioning earlier, I'm like some of these people that I met, like with these millennials, it's probably not a bad thing that we're not having as many as. But other than that, that's it.
Justin Shelley (31:44)
Not pumping out as many of them. Mario, you're so rude. So rude.
Brad Groux (31:49)
Get off my lawn. is the Get Off My
Lawn podcast.
Justin Shelley (31:51)
God damn. How old are you again, Mario? Jesus Christ. If you're just a baby and you're talking like that. Okay. So, here's, here's kind of a horrifying question I'm going to throw out because like I mentioned before, if, if this was taken from me at this point, and by the way, I full confession, I'm a late adopter.
Mario Zaki (31:54)
45 45
Justin Shelley (32:12)
It's not been very long that I've really been using AI to make my life and my business better. Um, but it has transformed almost every part of my life. So if you took it away, that would be a big problem. So on that note, you guys ever heard of the, uh, drug dealers who will run around to the playgrounds and the elementary schools and feed them a little crack and they do that. They give it for free. Like try this. This is a great stuff. And then what? Yeah. They've got lifelong.
Brad Groux (32:37)
Get to taste. Yep, you get to taste.
Justin Shelley (32:41)
And then now it's no longer free samples. that price goes up. So my fear, this is a real fear that I have. I'm paying 20, 25, 30 bucks a month. I don't even know. Actually, I've got a few different models. it's, it's each I'm paying about that much. How long before we're all hooked on the crack and this price goes through the goddamn roof. Because I will tell you the AI industry is losing money right now.
Brad Groux (33:03)
I again, think technology is a great equalizer. So open source models, you can buy a $10,000 Mac studio right now. I guess that's a big investment, but it's the $10,000 behemoth back Mac desktop. And it can run DeepSeek, the largest model in the world, completely in that own little black box. DeepSeek is an open source model. It's Chinese. I probably wouldn't use it for my US based business, but that's just as an example. And then the cell phones are getting smaller, smarter and smarter and smarter. so ⁓ Meta, which is Facebook, they have their llama.
which is an open source model. Google has Gemma, which is an open source model. OpenAI has their open source model that they're reaching. You're to be able to run those on your phone. The hardware is outpacing the need for the models, and the models are getting better and better and better, more efficient as well as we use these things. And so.
Justin Shelley (33:49)
Isn't it
interesting that we're back in that game of technology having to increase. Cause we, we plateaued for awhile where all the hardware, the processing, the memory, all that power was enough. And now all of a sudden we're back to this game where we have to catch up on the hardware side.
Brad Groux (34:03)
I think it's
Mario Zaki (34:04)
Brad,
have you talked to Apple? Because they're behind on the AI game on their phones. Their AI ⁓ is shit. I don't think it was even fully released.
Brad Groux (34:15)
Yeah, that's the beauty is though, ⁓ I don't need their model. can use their hardware to run. right now I'm on a Mac mini, I run models on that. I have a MacBook Pro, I'm actually a Microsoft guy, believe it or not. And I have a gaming desktop that I run bigger models on. It's democratization, I think is the key point. ⁓ Your data is what's valuable. And so if you're going, that's my key to businesses specifically. It's okay as like individuals, you're using ChatTPT, you're using Gemini in the cloud.
What you should be thinking long-term for your organization is how can you make sure that your data stays within the confines of your company and is what you do. so it's used. Yeah. Yeah.
Justin Shelley (34:51)
There we go. Thank you for that. We need to transition. Like I said, we're going to
a minute ago and dig into security.
Brad Groux (34:59)
Yep. Because, ⁓ you know, we see data breaches all the time and all these large organizations, these, these AI companies aren't going to be any different. There are the law, massive threat vectors, right? So why would they not want, want to be accessed? And so that's, let's be honest. They like open AI did some really unethical things to train their models. Everybody has, they scraped the open internet. Like how do they know how to tell you summarize Harry Potter for you? Well, they, they actually fed it Harry Potter, a copyrighted work, you know? So, ⁓
Just know your data is what makes you secret and all of the big organizations know this and that's what makes you that's your secret sauce and so you should try to do whatever you can to keep that within the confines of your organization. So have those conversations now is like we set that foundation. Now let's build that garden and all of our data stays in turn.
Mario Zaki (35:43)
But how does the, and I know, I know they're, know, whatever you put into it is out there, but how, like say I have a conversation with my AI and I say, I have a new business model that I want to do. it, you know, I'm thinking of doing this, this and this. How does that information then gets translated to somebody else? nobody else, you know, you actually have to ask it what new ideas people.
are entering today? that what, like how do they get that information?
Brad Groux (36:19)
Now how the models are trained is basically they strip all privacy information off. But if you have a business account for open AI, you have a little checkbox in your administrative panel for open AI business. And you can say, don't use my data for training. that means. Yeah.
Justin Shelley (36:30)
Which is by the way, the default and that's the difference. One of the
primary differences between a personal and a business account is that the personal defaults to give the world all your information and the business defaults ⁓ to keep it private. Now, how much do I trust that they're keeping it private? That's a whole different podcast episode.
Brad Groux (36:43)
Yep. that's what
Yeah, no. And so that's the thing is like you you have to have that level of trust ⁓ internally what we would tell people like use that for generalized information again, they're not going to be able to translate one to one like ⁓ someone input this and so now it's out there in the world. That's not how training a model really works. But just know that ⁓ look at it from a marketing perspective. Why look at LinkedIn right now. Why do all LinkedIn posts look similar because everybody's using chat to PT. And so they're trying to say the same sort of things and the same sort of models. And it's going to be unless they've given a context of like
your tone and your dialect and how you actually like to write, it all is gonna look very similar. You're see ⁓ dashes everywhere. That's like, like you see dash or emojis everywhere. It's probably a chat to pity.
Justin Shelley (37:25)
got to
tell GPT to, I guess if you yell at it enough, it'll quit doing that, but you have to tell like a hundred times in a hundred different ways to kill that dash or strip them out. We can copy and paste. I definitely do that. ⁓ and also what you just said is why I absolutely refuse to just keep posting these AI generated posts on LinkedIn. I love that we've got this where we're actually talking human to human as far as anybody knows. and I can take these clips and that's my, my LinkedIn content, right? It's, it's nobody.
Brad Groux (37:36)
Yep. Yep.
Justin Shelley (37:55)
will ever know that this is all fake and generated through JetGBT.
Brad Groux (38:00)
Yeah, I agree. think the human to human touch, the coffee meetups, the lunch meetups are going to be more more. Again, you're to have more time to do that anyways, because you've automated these menial tasks. You'll have more time to have those meaningful things. And I think we saw that during COVID, right? You know, people started garage businesses where there were hand-making leather goods and things like that. People were tired of throwaway society and stuff. We we realized like, crap, we have we're we we can't get masks because they're all made in China. Like, let's make our own masks. Let's start our own 3D printed pipeline. Like I think
Justin Shelley (38:04)
Yeah, yeah.
Right.
Mm-hmm.
Brad Groux (38:29)
technology is going to teach folks that they can, again, just think of all the great ideas we've had over the years where someone's had a great idea, but they weren't a developer. Like now they can actually, but they are someone who's a go getter who can like see, and you guys know it really well because you work in IT, I work in IT, I'm a jack of all trades. can code, but I wouldn't call myself a developer. Like I understand the basics of object-oriented programming. Now someone with an idea is like, and ideas are a dime a dozen. Now you're going to see who delivers on those ideas and uses these tools to help push out and do something great.
Justin Shelley (38:41)
Mm-hmm.
Mario Zaki (38:59)
Now, Brad, ⁓ without getting too deep in full quote, your services, usually how much would your services run for a small business to help automate some of this stuff?
Brad Groux (39:13)
So we'd like to start as little as like 10 hours, which would be about 2,000, 2,500 bucks. Like, hey, let's just talk about what your long-term goals are. And then from there, we can kind of go from there. Like automating an AP process for, let's say it's a company that has 150 invoices a month, you're looking at an initial cost of 10 or $15,000 and then a monthly recurring cost of licenses and actually processing the invoices stuff for 40 or 50 bucks. Like that's the thing is these models know 80 to 120 invoices is nothing compared to like...
from the model ingestion aspect of it. But it is everything whenever you have two or three people processing invoices and timesheets every single day. So the initial cost is probably going to be $10,000 $15,000 to automate. First, a small to mid-market company, their AP process, if it's documented wealth, not, we'll help you document what that actual process is. If you don't even have the basic understandings. ⁓ Our goal is always to meet you where you are too. So if you have someone really smart with your organization that you want to get involved,
Hey, we'll inject them into our work process to streamline that process too.
Mario Zaki (40:13)
Yeah, and $10,000, $15,000 at first you can say, you know, wow, that's a lot to get this done. you know, just like you said, if you have one, two, three people that, you know, you're paying at least $50,000 a year, you know, for these guys, now you could have them, you know, do something else, elevate their, you know, them to a different task or, you know, unfortunately kind of give them the snitch.
Brad Groux (40:26)
Mm-hmm.
Yeah, and like, again, another example is, again, we work with construction and engineering firms a lot. And so these, these project managers, program managers, the guys who oversee these multi-million dollar build outs, they have to go out and take pictures every week and provide progress supports and all these other things.
they can literally turn on Microsoft co-pilot on their phone and just as they're walking around the job site, it will record what they're actually looking at and what they're seeing. They can snap pictures with their phones. At the end of the day, they upload all of that to a single folder or whatever. And then you have an automated process that summarizes all that into a report and dumps it into a PDF format at the end of the week. And this is someone that's billable at 250 to $300 an hour to the customer. And if you can have them, he can only be on this job site, he can be on job sites more and be billable more, the cost is easy to prove.
I always tell people don't really focus on the ROI, focus on the time. Because again, you can do the basics of like, I saved that guy eight hours. We can bill him out at $250 an hour. Okay, you have what that is. It's two grand, so.
Mario Zaki (41:26)
Mm.
Justin Shelley (41:37)
All right, guys, we, I think we've said everything that needs to be said. then maybe then some, maybe, maybe a little too much sometimes. ⁓ we're going to wrap up. We're going to go with key takeaways. So Brad, at this point of the show, we just kind of go around the room and if, if your ideal listener heard only one thing out of your mouth, what would you want that to be? And I'm to start with Mario so that you have a little bit of time to think about that. and then Brad, you go ahead and tell us your, ⁓ your, your key takeaway from the show today.
Mario Zaki (42:04)
Yep. For me, I think it's something Brad mentioned earlier is that you have to sit down and see what you're doing on a regular basis, either daily or weekly and see if, you know, will it work if you could automate this to AI or to some odd, you know, type of automation and what you would do with your time. If you can get that one hour to three hours, four hours back from your day or week. I know.
There's a lot of other things I would love to do if I can get like four hours back from just my week.
Justin Shelley (42:40)
All right, Brad, what's your key takeaway?
Brad Groux (42:42)
I'd say if you're a smaller mid-sized business, you're probably already sitting on a goldmine of unused tools. So if you're Microsoft 365 or Google Workspace, you have automation tools already built into your platform that you're paying every month for for every user. We can help turn those into a strategic advantage. And so we can teach you the basics of automation. And trust me, once we show you what's possible, again, you know your business better than we ever will. And so you're the
the wheels are going to start turning and the light bulbs are start coming on and you're going to come up with new ways that we never would even have predicted. Even though maybe we've worked with people in the same industry because you think differently and you operate differently. And my, biggest takeaway is like, don't be afraid. I started a podcast called start small, think big. That's literally what you do. You know, you, how do you eat an elephant one bite at a time? You know, it's not a marathon on a sprint. can use any, any number of sayings to help get that a point, but like we're, all still learning. And like Justin said, he started slowly. That's
Are you saying he's behind? But no, he's not really because you know why we're like toddlers in the AI space. The age of the Internet, we're AOL. And just see where we've come with the Internet. And so we have the ability. Now is not too late. Five or 10 years down the line may be too late and your business may stroke.
Justin Shelley (43:39)
Mm-hmm.
Mario Zaki (43:42)
Okay.
Justin Shelley (43:42)
Yep. Yep.
Yeah. And mine. And so this has just been the thing that has been the most shocking that I've learned more than anything else in my journey with AI is find the place where you feel stuck, where you think you're just, you can't get past it and hit AI and just start talking to it like a human and, figure out how to solve that problem. ⁓ I think in, know, in life and business, we all have these areas where we're just like, you give up on it almost. It's like, this is too hard of a problem for me to solve.
I'm going to go do stuff that's easy. Find that thing that's really hard that's holding you back and get it solved. We have resources now that have never been available to the human in the history of humanity. Get after it and start using it. So, all right guys, great, great show. Call Brad. That's, that's right. Brad, go ahead and give us your, your website. How do we get people ahold of you?
Mario Zaki (44:33)
Yeah, and if you can't solve it call Brad
Brad Groux (44:44)
So the easy way to find me is to search my name, Brad, B-R-E-D-G-R-O-U-X, that's Brad Gru. I'm the only one in the world, so if you search it, it's me. Yes, there's embarrassing things from 25 years ago on like sports forums and things, ⁓ and video gaming. ⁓ But yeah, and our website's digitalmeld.io. ⁓ Digital meld is a play on digital convergence. Bill Gates talked about that in like the early 90s. our goal is to help you again, punch above your weight class. And really, we love being evangelists. I love to just talk to folks. If you don't have the...
Mario Zaki (44:53)
Ha
Brad Groux (45:11)
the budget for it. Hey, we can still hop on a 30 minute call and I can give you some guidance. Like that's again, the way I look at it, you put good things out there, good things that are come back and the more people that learn about AI, my 72 year old mother uses AI to help balance the books for my company through with QuickBooks. So if she can do it, if she can do it, you can do it. You know, like really it is, is, it is groundbreaking. You know, how many seniors now use smartphones? Do you see, and like you would never thought that they would be and they live on Facebook. They live on to, they keep track of folks. You can do this. It's, it's
Justin Shelley (45:15)
I love that.
I love it. No kidding.
Brad Groux (45:40)
Technology is a great equalizer. Always remember that.
Justin Shelley (45:43)
love it. All right, guys to listening audience. If you're on Spotify or Apple or whatever, go to the show notes or the summary or whatever, and we have clickable links in there. There will be a link to Brad to your website and a note to search the one the only the great Brad grew. So Brad, thank you for joining us, everybody go to unhacked.live. If you want that free assessment from Mario or myself or Brian, even though he's not here, we'll give him a plug. And we'll do a free cybersecurity audit.
Brad Groux (45:59)
Thank you.
Justin Shelley (46:10)
And I love the way you said that Brad, I'll throw my hat in ring there too. If you just want to have a conversation like let's learn together, let's grow together. Let's build a community around this around us. AI, cybersecurity, technology, or just life. You know, this is, this is what we do. This is how we help each other out. So, I do believe that it is all about community as technology continues to grow and develop. ⁓ we are allowed to be a little bit more human. That's, that's my take on it. So unhacked.live for today's show.
future shows, links to all the platforms, social media, et cetera. And just go to your show notes and you'll get a link right to Brad's contact information as well. So guys, that's it for this week's episode of Unhacked. Let's say our goodbyes. Mario.
Mario Zaki (46:55)
All right, thank you guys. And just remember if you're staying up at night, worried about your business, let us help you so you can get some sleep at night.
Justin Shelley (47:04)
All right, Brad.
Brad Groux (47:05)
Yeah, and if you want to know about AI, started a podcast called Start Small Think Big where we talk about AI and automation and we start at that foundational level. We're about 15 or 16 episodes in. Now we're talking to experts that kind of echo what we do. And so look for that on your favorite podcast platform.
Justin Shelley (47:17)
making a quick note to put that in the show notes as well. So you can just click it and go right to Brad's podcast. So again, Brad, thank you for being here. Mario as always, and I am Justin. Remember, listen in, take action, and keep your business unhacked. See you next week.
Brad Groux (47:33)
Bye.
Mario Zaki (47:33)
Bye guys.
Creators and Guests

