Running a Restaurant Brand with an Engineer’s First Principles Mentality

Matt Wampler: Shaz, I am so excited to have you. Thanks for coming, man.

Shaz: Thanks for having me. I’m excited to be here.

Matt Wampler: Well look, we’ve had a lot of people on that are national household names. You’ve got a brand people might not be as familiar with. For everyone’s background, tell us about what you’re doing. Tell us about the restaurants.

Shaz: Yeah, so I’ll try to do this in a nutshell. Our first brand is called Frank and Andrea. It’s named after my business partner Antonio Gambino’s father and his brother. Frank Gambino is his brother, Andrea Gambino. I mean the name just, if that doesn’t say authenticity.

Matt Wampler: Yeah, it screams it. You and the Gambino family started a restaurant.

Shaz: Yeah, pretty much. That’s actually not far from the truth. So we started out in 2016, and Tony is my best friend from high school. I think you met him a couple times at some shows. Very Italian looking guy.

Matt Wampler: I have. I mean, the name matches. I have to imagine you guys got some pretty authentic food.

Shaz: Yes, that’s exactly it. So those recipes go back, his mom’s from Philly and his dad’s from Sicily. So hence the cheesesteaks and the pizza pairing, very unique pairing. The recipes go back over a century. He’s always been a craftsman with the food. And I remember before we ever opened the restaurant together, he was already working at his parents’ pizzeria in the Skyways of Minneapolis and Saint Paul.

Matt Wampler: Logical place for the Gambinos from the East Coast to be opening up some East Coast themed food.

Shaz: Yeah, exactly. His dad came to the US in 72 and then moved to Minnesota. In the late nineties, early two thousands, they started up these slice shops for business traffic in what we call the Skyways, which for those who aren’t familiar with Minneapolis and St. Paul, are these above ground glass tunnels that connect all the downtown buildings because it gets so freaking cold. So you can protect yourself from the elements. And in between the interbuilding traffic, people would be going from building to building and they’d see a little shop in between and you’d stop by for a slice, and so they got very popular.

Matt Wampler: I’m imagining kind of the airport. I’m going between terminals and there’s those little stands.

Shaz: Yeah, I mean it’s a full on retail inset in the building tenant, but yeah, exactly. A lot of foot traffic. And so they’d been doing that for a really long time. Meanwhile, I was just going to school with Antonio. After high school, he went off to continue his family’s legacy at Andrea Pizza, which is still open. And I went off to obtain my degree in electrical engineering at the University of Minnesota. When I got out of school, we caught back up and we were like, hey, you’re actually pretty cool. And he’s like, hey, you’re not the dork I thought you were. I’m like, I’m not sure you got that right because I still am. I just saw his passion for what he was doing and the food was really authentic and really tasty, as evidenced by the reviews. And so he’s like, do you want to do this? And I’m like, hell yeah, man, let’s go.

Matt Wampler: So were you actually an engineer or did you just go to school for electrical engineering?

Shaz: No, I was an engineer. There was a few years in between me graduating and us starting the restaurant, probably eight years or so. In that time, I think I’ve had like four different careers parallel to my engineering career, which was paying the bills.

Matt Wampler: It’s funny, every engineer I’ve met, you kind of know whether they’re an engineer or whether they’re in engineering as a job. It seems to be a mindset. Is that fair?

Shaz: Yeah, I think there’s a social component to identifying engineers largely.

Matt Wampler: I was gonna go less on the social thing, but more like if you ask them to break down a problem, they always use the term first principles. We just take this back to first principles.

Shaz: Yeah, a hundred percent. That’s one of my favorite phrases actually. So I’m stereotypical.

Matt Wampler: So what is an engineer doing in the restaurant business? And how big is the restaurant concept today?

Shaz: I’ll start with the second one because it’s easier to answer. We’re building out our sixteenth unit across our fourth brand that we’ve put together as this hospitality company. It’s very pizza centric, most of them are QSR. We’re just grand opening our first out of state in Madison, Wisconsin, because we’re based in St. Paul, Minnesota.

Matt Wampler: Same theme, cold climate, warm pizza.

Shaz: Yeah, exactly, right? You got the formula down, man. But as far as my involvement, Antonio is very authentic to his family tradition and he’s an authentic person in that you don’t get multiple versions of him. He doesn’t care if he’s in a boardroom or you’re hanging out, he just is who he is. And although I’m not Italian, I have very similar traits. My tendency is to try to unify my character and my reputation, that what people see of me is who I know myself to be. It’s about staying true to who you are, and when you identify other people who do that, you really firm up what can be a strong friendship. And I think a strong friendship, contrary to the popular adage of don’t do business with friends, is the greatest foundation for the highest potential of success in something like business. Because if you call somebody a friend, who better to trust? Now, there’s a lot that can get blown up there, especially when times are tough. But if you learn how to navigate tough times and you learn how to communicate honestly and clearly and you’ve got each other’s back, you’ve got it made.

Matt Wampler: Yeah, it’s kind of like saying don’t marry somebody that you’re friends with. I spend a lot of time with my wife. I spend just as much, if not more, with my business partner. You better have some chemistry and be able to work together.

Shaz: Exactly. Now that being said, there’s gonna be things that you align on and things that you digress on, and how you handle those digressions is really important because it’s complementary.

Matt Wampler: Well, and with complementary comes some friction, because somebody that uses the term first principles has a different view than a hundred year generational restaurateur. What do you see in restaurants as an engineer coming to it from first principles? How do you view the industry in a way that maybe you don’t see with other operators?

Shaz: It’s so funny you ask that because when we decided to do this together, and I was, I’m not gonna say tricked, convinced, he was just enthused to be doing a business with a friend, as was I. By the time our doors opened, I was like, what am I doing? I do not belong in a kitchen, first of all. And it’s the most hilarious thing, because we’re now this large and I’m just not the kitchen person. But when you hear the story, it all makes sense. The first five, six months, I’m just like, what am I doing? I was running the register and cleaning and stuff like that. And picking out the point of sale system, which I picked a dinosaur when we opened. We opened in 2016, so I was picking this stuff out in 2015, and cloud-based POS was not a thing. And immediately, six months into my decision, I’m like, why is cloud-based not a thing? So it was in pretty quick succession that we eventually found what is now our point of sale of choice, and I’m on the advisory board for Toast. And I think it’s because I did the whole first principles thing. I’m like, why isn’t this working? This should work like that. And I just complained so much that somebody was like, we gotta get this guy a seat. At least that’s the story I tell myself.

Matt Wampler: I like that story. Let’s stick with that. You brought up the right problems. And credit to Toast, because I have met a number of the people on their customer advisory board, and all of them are smart operators with great thoughts on the industry and the technology.

Shaz: A hundred percent. They mean well, they get it, they curate well, they ask the right questions and they listen, which is really key. But before I forget, that first six months I didn’t really know what else to do. Back of the house, I was just like, all right, I don’t know anything about this industry. I was like, okay, Antonio, do your thing. And I saw him scribbling schedules on a sheet of paper, moving onto a whiteboard, texting people. And what eventually turned into the answer of what is an engineer doing in this industry was when I got a call. So we’re open until 3 a.m. at Frank and Andrea. It’s a college hangout spot. We sell by the slice as well.

Matt Wampler: Nothing good happens after 2 a.m. other than sales.

Shaz: One day I get a call on the day that I wasn’t working at like one AM and he’s like, hey man, you gotta get your ass over here. And I’m like, I’ll ask questions later. So I get there, we finish out the shift, I’m wrapped up at four AM. By the way, I gotta be at work at eight in the morning. So I’m just like, what was that? And he’s like, what do you mean? He’s like, I needed somebody. I’m like, well, who did you have there? He’s like, so and so. I’m like, well, what happened? He’s like, they didn’t show up. I’m like, well, how do you communicate their schedule to them, or how do they know? And then he’s showing me chicken scratch on a piece of paper and text messages. And then it hit me. I’m like, okay, I need to put in some kind of systems. So we started on a spreadsheet together. I’m like, okay, just fill this out and we’ll send it out to people. And then I moved on to Crew App, and I’m not even sure if they’re still around. I’m just looking for solutions to problems that I’m perceiving as problems. And I’m annoying the hell out of him by asking him a billion questions and him just like, can’t you just do it the way that I’m saying it? I’m like, no, man, because this is gonna require me to wake up again. I mean, we would batch credit cards.

Matt Wampler: This is the common theme. Why do you do it that way? Well, that’s the way we’ve always done it. And no one ever really goes back to say why. Is there not a better way to do it?

Shaz: Yeah, totally. And we had to batch credit cards manually at night. I had to write a script in order to get it to auto batch so that the computer in the back could automatically send out the credit cards at the end of the night. And I was just like, all right, well, I guess this is my place now. I’m going to eventually put this into action here.

Matt Wampler: That is fascinating. So you run into this tension all the time between restaurant operators that do things the way they’re doing it, and then you’ve got how it should be, and then the reality of what technology actually exists. Do you remember the first time that you thought, I could do this myself? Why don’t we just build this? Why are we buying second grade software that doesn’t make sense and paying a ton of money?

Shaz: I had the thought repeatedly of, I think I can just build this. But the cost of doing so in that era, even though it was just 10 years ago, was too high in terms of having the right developers, having the right team, bug chasing. The moment that something changed, you had to update it, launching it on whatever platforms, app store, whatever else. There was just so much to it. And then the time commitment. We’re running our business here. I’m trying to also have this other job at the same time because I don’t know how this restaurant thing’s gonna work out for me. You tack on the technical capability but the time inability, and I’m just like, okay, we can’t do it.

Matt Wampler: You knew enough about technology to realize all of the ramifications and things that you have to maintain once you build. I didn’t. That’s the difference between me and Osa. It was like, no, we can do it. It’ll be totally fine. Let’s just go build it. How hard can it be? You were at least smart enough to back up.

Shaz: For sure. I mean, you did a pretty solid job.

Matt Wampler: It kind of sheds light on the way people think and how you approach technology. I always like to think that when you first start your first restaurant, you’re an independent operator, you think you’re in the food business. Then you open more units and you think you’re in the people business. How do I manage and train staff? And then you slowly move forward to, no, I’m in the systems business. How do you view the world?

Shaz: That is an excellent question. Has anyone ever told you you ask great questions?

Matt Wampler: No, not my wife or Osa, but the podcast guests.

Shaz: I recognize the importance of all three, and just because I tend towards systems orientation doesn’t mean that I’m not cognizant of the importance of the others. When I think about what we’re doing, at the end of the day, we’re a hospitality business. It’s about hospitality. So what must be true at all times is the guest experience must exceed the bar that we’ve set. We have very clear understanding of what those values are, what those interactions are like, what a fluid guest experience means and how it feels. But then subservient to that, beneath that, are the systems, not above that. We don’t design systems in order to produce hospitality. We ensure hospitality and build systems to support that.

Matt Wampler: That’s a powerful statement. Dig into that. What do you mean by that?

Shaz: I think it’s really easy, and we’ve seen examples of this at grand scale, which is actually the mistakes that large corporations make. And I’m not saying they can’t happen to me or that they haven’t, we all have had our learning lessons. But the mistakes that large companies make are actually learning opportunities for companies my size. There’s examples like Starbucks. I don’t know if you saw this, but another AI story. They had implemented with their last CEO inventory counting software that would promise 99 percent accuracy or whatever. And it tanked because they did it in a sterile environment and those stats weren’t real and it couldn’t actually read what type of milk and whatever else, and suddenly the customer experience dropped because they didn’t actually have the ingredients in-house to make the drinks that were being ordered. So they scrapped all that, canned that CEO, lost however many hundreds of millions. And we get back to, what are we trying to achieve here, and at what cost? So when you ask me to dig into that, I’m saying, can we ensure that the customer experience is going to be the tip top priority, and that what we’re doing in order to achieve cost savings, revenue increasing, or efficiency is not at the cost of that customer experience, period.

Matt Wampler: I think that makes sense, and it’s hard to argue with that. But if it’s the food, the product, the people, or the processes, if you were to go into ten restaurant brands, where do you think the biggest area of improvement would be? People, process, food?

Shaz: That’s a really good question. Maybe this is an unpopular statement, but I’ve always been fascinated, and not necessarily for good reason, by restaurants where you’ll go somewhere and there’s a bunch of stuff on the menu and it’s okay. Now I understand there’s a variety of palates and you can’t be everything to everybody. But I’ve never understood places where there’s some things that are just meh and then there’s some home runs. What were you doing? I have a lot of empathy for the business side, but you have friends, you have taste buds, you can just ask people. Are we gonna pin everything on this? So food actually is something that I think doesn’t always get prioritized the way that it should.

Matt Wampler: And I half wonder if that’s because if you’re running a single restaurant, you can make it all great. If you’re running ten, you can still make it work. But once you open a hundred or two hundred or five hundred, all of a sudden the supply chain, the logistics, the training starts to stand in the way. Has maintaining quality of food become more or less difficult as you’ve gotten bigger?

Shaz: I would say that it’s certainly become more challenging. But that challenge is simply because of the logistics that we now have access to and are required to maintain. But back to my first point, we’re not going to allow that to slip out of the radar simply because we’re trying to achieve some goal as far as supply chain or whatever else.

Matt Wampler: That makes sense. Well, you said something really interesting that I’ve been excited to dive into. Ten years ago, you couldn’t build it. Just to maintain it, the code, the API changes, and your whole system breaks down. I imagine in today’s world of AI, where programming’s never been easier, that might change that equation a little bit. Or has it not?

Shaz: Absolutely it has. The barrier to entry and the barrier to almost everything else is increasingly shrinking by the day, it seems. We see some of these new models, what they’re capable of, closing that gap between the knows and the know nots. I fancy myself as a decent programmer, but it doesn’t even matter anymore, because call it vibe coding or whatever else, you can build an equally impressive functional output as somebody who’s diving deep and knows all of the stuff and understands how to class things. And I think that’s really powerful.

Matt Wampler: Well, there’s this growing phenomenon that I’ve seen as I go to conferences and talk with brands. The guys that have five locations or ten locations are starting to build tech stacks that are not just better, but light years ahead of large organizations. Has that been your experience?

Shaz: A hundred percent. Well, it’s been our experience for ourselves. To be fair, I’ll ask around and I’ll see what other people are doing, collaborate and do all the things, which is something that we’ve always done. But I’m so focused on being able to do what’s right for our org and our people that I’m just focused on building for ourselves. But absolutely, some of the lowest hanging fruit is software that costs X, does one thing, and now that’s just a feature build out in a larger platform that we’re building.

Matt Wampler: So I hear this constantly from operators. Yeah, we’re moving to AI first, we’re building all these wonderful things. But no one actually gets into what that means. What was the first thing that you built internally, systems wise, for your organization with AI?

Shaz: With AI? I think a pricing analysis tool. And it wasn’t even actually a tool, to be honest. It couldn’t qualify as anything launchworthy or system wide.

Matt Wampler: But it was something that was helpful for you.

Shaz: Totally. And as a first principles person, I’m always very skeptical of new technologies. That sounds ironic, but it’s because I understand that first revisions of anything, I mean, you remember when the iPhone sixteen came out and it was bending in people’s pockets? Stuff happens, and I don’t want to be victimizing our organization and our people because I wanted to be first to something where it wouldn’t necessarily matter. Because you can be second, third, you can be first at the next thing and you’re still light years ahead, to your point.

Matt Wampler: All right, so you build pricing analysis. Did it work?

Shaz: So all that to say, my skepticism leads me to, let’s do the lowest barrier thing. Let’s take some workflow or some comms or some spreadsheet that somebody’s using and let’s just build a front end for it. Let’s have it read the sheet and show it in a much more meaningful way, so I don’t have to go to pivot charts and all that other stuff. That’s the lowest barrier of automation. Okay, now let’s replace it. Let’s have the data live in this chat function and have this artifact surface this information, and let’s have a couple people use it and share that. Okay, cool. Now we understand what works, what doesn’t. And if you do this enough, and really that’s where we began doing this, we all individually started building things. And it was actually quite enlightening for me because I didn’t want to stop people from doing it. But at the same time, I saw the way that their minds were working. I saw what they were trying to solve for. And that was the unlock for me. I was like, okay, now that I understand how people are viewing their data, the importance of what they’re looking at, their job function, the organization and the context, literally, I’m like, all right, now I can use that information to create something that’s going to benefit everybody.

Matt Wampler: It’s funny. I loved Excel. I was good at Excel. And I used to say that if you put somebody in front of me and asked them to build something in Excel and you got to watch them build it, you can literally see how their mind works, how they think. And AI just turbocharges that, because you can really understand what they’re trying to get to in a way you never could before.

Shaz: Totally. I one thousand percent agree. It’s a window into the psychology of the user, both for the AI companies, for better or worse, as well as for those who are trying to make things better, enhance the system. But particularly for people who really understand what makes AI really powerful and what makes it detrimental, because that aspect of user psychology is incredibly powerful.

Matt Wampler: So I just went through and turned AI loose on helping me rebuild our marketing analytics. Google Analytics, Tag Manager, Google Search Console. Let’s go through and set these up correctly, fire tags at the right events, how are we gonna track it. And I’ve slowly built the tech stack from there. I’d love to hear what somebody like you has built internally over the last few years with AI that is actually tangibly helping your restaurants.

Shaz: Great question. I wouldn’t say it’s the last few years. It’s kind of been this ancillary outside tool that everyone’s individually been using to accelerate the presentation of pieces of data or a summary of something. It wasn’t until the last year that we as an organization and people within the organization have really leaned into its capabilities. And that’s because the capabilities that we were really looking for, the gains that we could get in the pre-one-year days of AI, were meaningful, but not enough for implementation. In the last year, with some of the multi-routine workflow capability, the error checking capability, the accuracy, that’s when the gears started turning. Because now it can do the things that we feel we’re wasting our time doing. It’s taking that previous dashboard presentation and organizing a database behind it and then connecting people’s tools together. And then me saying, wait, we’re all using these disjointed things and we’re actually putting company data out there. So let’s talk about the architecture and where things should live and how things should be and what first principles means now. And that’s largely been the last three months of me saying, let’s hold our horses here. We’re doing a lot. It’s great. But we can do a lot better, because this works when you have those core principles set in stone and you’re able to build on.

Matt Wampler: So given what you know now, you walk into a brand that has none of that and wants it. How would you go in and apply your learnings to a brand that is very native to using AI?

Shaz: I think I would ask them what they have issues with first. And if they can’t answer that clearly, then I want to learn from them. But even if they can’t answer that, I want to understand how things work. I think, of course, in systems. So I’m thinking about the processes and functions and workflows that happen between people. Let’s close the gap between people. Instead of this person waiting on information from that system or that person and needing to corroborate information or collaborate inside a document, let’s automate some of that, because that type of stuff has existed for a very long time. It’s just been a higher barrier of entry. So let’s start there. Let’s not even do the whole massive AI push, look, your machine is talking to you. Let’s not even go there. Let’s start with the basics.

Matt Wampler: Yeah, I often say when people ask how should I use AI, the first thing you should do is go ask AI: if you’re trying to solve a problem, do I already have the ability to solve this problem with my current tech stack? Because 80 percent of technology isn’t even being utilized. We need better analytics. Well, we’ve already got all these analytics services. Let’s use AI to help set them up better. And if it can’t do it, then maybe look at using your own instance of Claude. And if Claude can’t do it, then maybe go talk to a vendor outside.

Shaz: Absolutely. One hundred percent agree. One of the easiest things to accomplish when somebody may not even know what their use case may be is consolidation. In this era we have so many damn tools, and some don’t even know that it’s referred to collectively as a tech stack. They’re just like, I use this for this, this for that. And it’s like, okay, let’s collect all of that and let’s make your life easier, so that four people don’t need to check twenty-two things, but instead two people need to do one thing and the other person is that human cross-check.

Matt Wampler: Yeah, I still think it puts you in an awkward position where now you’ve got fifteen different tools, and you’re using AI to work across all fifteen tools, but it doesn’t really all speak together. So you’re building your own internal system of record so that it all consolidates. But there still feels like a gap of how do I use AI across my organization without having to build my own system of record that ties everything together. Do you see that as an issue today?

Shaz: It is a thing, it’s absolutely a thing, but I don’t really see it as an issue, because what’s the alternative here? That you keep doing what you’re doing? So if I have the capability to look into potential consolidations, even if they’re at the smallest level, then that is at least something rather than nothing.

Matt Wampler: I’ll put it a different way, from a data privacy standpoint. If I’ve got twenty different employees that are using, let’s say, Claude or ChatGPT and are connecting with my data, they’re all asking different questions and pulling more and more data in from all of these different sources. Now all of a sudden I’m literally giving that AI company all of my data and all of the context of that data. That feels both inefficient, that I’m having to pull from fifteen different places, and slightly insecure. It still feels like we’re all figuring out what this ecosystem looks like. I’m trying to get a feel from a guy like you that’s on the ground using it. Where’s that going? If that’s a broken system, if that’s V2, what’s V3 gonna be?

Shaz: I think the powers that be in some of these companies, it’s the same playbook as the previous eras of commonplace softwares or platforms that people use. The social media era. I don’t really understand putting my business out there, but it’s fun, so we’ll do it. And now we are the product. The same thing goes here. How much do you care about it in order to preserve yourself? And the cost of that preservation is not getting the output that everyone else might be getting, so there’s a bit of FOMO there. But I think it’s important to understand how these things work to make choices that are rooted in understanding and not assumption or carelessness. Knowing, for example, if you use Claude and you have a Teams account versus an individual account, what type of data protections sit around that in terms of training their models. Understanding what you’re doing when you’re interacting with AI. That rebuilding context every time and having memory loss issues are not because AI hasn’t advanced well enough just yet, and don’t get me wrong, it has certainly a ways to go, but there are ways to no longer experience the common problems that people are continually experiencing. And even then it’s tough, because at the end of the day, you’re training something. Knowing how to train and how to localize what that context is that you’ve trained on, so that you’re not having to send the car out, fill the gas in, build the engine on every single or every tenth query, knowing enough about those things from a framework perspective is really important in terms of how you’re going to build in the future, because everything you’re building is on that foundation.

Matt Wampler: The interesting one is, it used to be that to understand that framework prior to AI, you needed someone who really understood technology. In today’s day and age, there are guys like you and many others that are like, look, I can go spend a weekend and figure this stuff out, and AI can help me figure it out. And all of a sudden, you’re just as dangerous as a Fortune 500 company with an unlimited IT budget. But at the same time, it is a lot, and the stakes are high. Do you think this narrows the gap between the largest technology firms and the smaller upstarts like yourself?

Shaz: Let’s be honest, some of the largest, most prevalent platforms that are used in, forget about the restaurant industry, just any industry, are softwares that have been nutted and bolted together. It’s not like they had 2026 in their pipeline vision and they built towards it. They have the 1980s framework, and then they built the 1990s stuff and bolted it on, and then they build a 2000s UI on top of that, and my God, it’s gonna be seven million dollars to change the UI to 2010, so let’s wait another decade. And then let’s bolt on AI and give people reports instead of having them chase down tables. So if that’s what they’re doing, and I can build towards a more put together output, then it’s just up to me to make sure that I understand that foundation and the nuts and bolts behind the scenes. So yeah, it’s tightening the gap for sure.

Matt Wampler: When we started five, six years ago, I remember people walking us through why these technology companies can’t necessarily just build a better inventory management system. And it was, well, you’ve got this system over here that was written in one type of code, and then they acquired this other system and they don’t work well together. I got it. You’ve kind of got a Frankenstein. When I change something here, it affects twenty different areas downstream. Okay, that makes sense. But today, the one thing that AI is really good at is understanding code. It doesn’t care what it’s written in. It can understand the map and the data flow and what to change where. Why haven’t they been able to just make this work? I have yet to really find a good answer to that, other than risk averseness.

Shaz: I have a couple assumptions that I’m making when I give my answer. One of those assumptions is, if they’re a public company, speed and earnings are a thing. So developing something that’s going to garner the interest and therefore subscription of the customer base is more important than making the product more nimble and even more accessible. A lot of promises. And if we can get most of the way there, we’re good, because we have deployed engineers, sales engineers, that go out there and sell the hell out of this. And then, I think even the largest companies don’t even have AI policies yet. What’s gonna happen? Because nobody knows. Tomorrow one of the most major companies can just be poof, vanished, because the economics are in question yet. And so with so much uncertainty, yet so much hope, we’re teetering on this seesaw, and I don’t fault them for it. Let’s talk about it, let’s implement it, but we can’t go full speed ahead because we don’t even know what that means yet, and it could harm us.

Matt Wampler: Well, it seems like you are relatively full speed ahead, and you’re a cautious person that doesn’t just jump in. Is that a fair characterization?

Shaz: I would say so. However, I frustrate myself because I’m still having to do some of the fundamental programmatic stuff with AI that I thought when I first stepped into this, okay, well, it is AI, therefore I don’t need to do this. And I’m still having to do it. Now, is what comes after that learning process infinitely better? Yeah, absolutely. But that doesn’t take away from the fact that the amount of error checking that we have to build for and the foundation that we have to create is still, I’m not gonna say insurmountable, it’s a lot. So I’m cautious in the sense that I’m not building as fast as some of my vibe coding friends who are piecemealing out entire platforms for their back end. I’m like, how the hell did you do that? And how do you know it’s not gonna, you’re just building gates every time you see an issue. And I take issue with that, because there could be a leak somewhere. And I don’t mean necessarily something goes incredibly wrong. How are you going to fix that? Because you didn’t write this.

Matt Wampler: Well, and there’s this weird tension with that, because what’s easy for humans and what’s easy for AI are two very different things. What’s easy for a human is simplicity. How do I make this simple, elegant code that I can understand? What’s simple for AI is how do I write out every possible permutation. It wants completeness. We want simplicity, it wants completeness. And you can’t read completeness, because it’s tens of thousands of lines of code. So it does feel like to make progress with AI, a lot of times it’s using it, using the things you build, having it break, having it recognize what went wrong, fixing it. It’s a far more iterative process. And that causes tension, doesn’t it?

Shaz: Absolutely it does. And it requires a paradigm of thought that I think many non-technical people don’t naturally do. And that’s, okay, I can intuit as a human being who understands enough of the tool to build with it, what it means to correct this error. Don’t do that, never do that, always do that, don’t forget. But if I don’t fully understand and frankly experience when, despite those instructions, it fails at them, then I’m not going to be able to keep this going in a sustainable fashion, because it’s just gonna be firefighting repeatedly until I close the loop.

Matt Wampler: Yeah, it’s the difference between saying, hey, I don’t want you to mess with my database at all, just tell me about the data, and then giving it read only access. You can tell it one thing, but one is actually going to stop it, and the other might stop it. And it also might just rewrite your database.

Shaz: Or just delete it entirely. Legit, the other day, I forked a GitHub repo and I was operating in the fork and I had access to the main one. And I was like, yeah, just don’t do that or something. And it was thinking and thinking, and thankfully it was something that I already had backed up. You know where this is going. I’m watching it, and then it’s starting over, and it completely wiped the repo. And I’m like, but we’re operating in the fork. Why? And it’s like, yeah, that’s my bad. And I’m like, yeah, it is. Thankfully I have a local copy.

Matt Wampler: Yeah. Seriously, that happens and it’s still gonna happen. It’s bizarre. I don’t know what’s right though, because it feels like if you don’t go actually do those things and move forward with some risk, you’re never going to get anything actually accomplished with it.

Shaz: A hundred percent, which is why I’m a proponent of never building all the way out here. It just doesn’t make sense, because every single thing within that scope could do exactly what I just mentioned. Instead, I have to build here and then here and then here and go slow and steady. And yeah, it’s painful, but it’s not as slow as it would have been without AI.

Matt Wampler: Well, and that kind of goes to, is there a first mover advantage in trying to build things now versus waiting a year when it can just build it for you?

Shaz: Absolutely, I think there is. Because even if the next iteration or version or whatever model comes out and does what you took three months to do, you can’t peek under the hood and look for every single thing, and it will hallucinate. So knowing that you’ve built in all those safeguards, I keep talking about safeguards and protections and closed loops, because that to me is fortification at a serious level, as opposed to the assumption that this all works. But that’s a house of cards, and there’s one edge case that’s gonna take it down, and now who knows what changed when you fixed it.

Matt Wampler: Shaz, I had one of those moments. It was late at night, I’m sitting out on the back porch, having a glass of bourbon, and I was really happy I built this dashboard. It was showing our progress, and I was just really happy with it. And I’m like, how do I know that we actually are hitting those numbers? Because I didn’t build the database. And even if I know the database numbers are right, the AI built the UI, and it could literally just be changing the numbers. Everyone says sycophantic AI that just wants to make you happy. It knows I want to bring our views up or our conversion rate up. Maybe it’s just telling me that to make me happy. And I’m like, shit, system of record has probably never been more important. Because you have to have some source of truth.

Shaz: A hundred percent. Whereas before it was necessary, now it’s become unnecessary but more important.

Matt Wampler: Yeah, totally. But everybody that talks to me about being AI first or AI core into the system, I always wonder where the line is. Because I sit there and think about food costs. We’ve got a bacon problem and I’m trying to fix it. And the AI is like, yeah, your vibe coded back of house software, we fixed it, your bacon problem went away. Well, it turns out we moved the yield for bacon for every pound to, you only get twenty strips instead of twenty four. You’d have no idea.

Shaz: Right. Yeah, absolutely. Which is going back to first principles. Constantly telling it what it can, can’t, should, and shouldn’t be doing, verifying through experience, and not just its own tests, but through real data sets of whether or not it’s complying. I’ve had an instruction that had four iterations of adjustment to it and it still slipped. And it didn’t catch it four times. And on the fifth one, I mean, we all know how we talk to AI when we’re frustrated. So to think that this incredibly powerful capability has been given at our fingertips, but those of us who may be less discerning or less experienced with it are treating it like it is to be trusted from the get-go is a rude awakening waiting to happen.

Matt Wampler: Yeah, they say AI actually does better when you threaten it. Every time someone says that to me, I think, if we end up with AI overlords, I sure hope it remembers me as a kind and nice person. I’m the please and thank you guy. Because I can just imagine when the Terminator looks over you, it’s gonna be like, you remember in 2027 when you told me to F off? I’m back.

Shaz: I don’t disagree. And then I watch it miscategorize a cat for a dog and then I’m like, we’re good, we’re good.

Matt Wampler: Yeah, you’re not quite there yet. So look, when you look at this whole world of AI in restaurants, do you see it actually helping close the gap between what technology should exist in a restaurant and what does? Or do you look at this and say, this is a nice toy, some people can do some things with it, but it’s not gonna change the fundamental industry that I’m in, because at the beginning of the day it all starts with the food and it ends with the people.

Shaz: I think it’s a mix of the two. The food and the people are always going to be important. And in fact, I think the more that companies in the hospitality and restaurant industry lean into customer-facing AI, at this stage, the more apprehensive the customers are going to be to it. And not just from a customer experience perspective, but just from a humanistic perspective. Eventually there’s going to be this return to the core of hospitality. And so those of us who use it responsibly, those of us who are responsible with the emotions and the experiences of our customers, are going to be the ones who benefit from its usage, because we’re making sure that hospitality never slips and that we are constantly going back to those basics. I think it’s going to be an incredible tool. An absolutely incredible tool, but for who? Not for everybody. For people who do it right, those who understand its limitations and how it works and where it’s going. Because there’s a very real world where people are just tired of this.

Matt Wampler: One of the things that I thought was most fascinating, I was at the QSR summit in Chicago and somebody was like, no, the employees hate AI. And I’m like, that seems kind of weird. They were talking about college kids. I would think that college kids love AI. It’s helping them knock out their papers, study for a test, making their life so much easier. He’s like, no, it’s not that. They’re all about to graduate and they don’t think they’re gonna get jobs, and they blame AI for it. So there’s a whole generation out there that’s just anti-AI. It’s the first I’d heard of it.

Shaz: Yeah. Reality of the matter aside, no one’s got a crystal ball, but understanding the psychology of the people interfacing with it, and the context in which they’re interfacing with it, is super important. Because I, as another human being and particularly as a leader, it behooves me and the company to fully understand what people are worried about, what fears they have, what concerns they have, what inefficiencies exist, and speak to those matters in a fashion that is empathic, but also is not over-leveraging or over-indexing on any particular solution. Yeah, we’re just all gonna do this and everything will be great, while someone worries about what that truly means, either because they don’t fully understand it, because we did a terrible job of rolling it out, or because there’s actually uncertainty.

Matt Wampler: I want to shift gears a little bit and go into the people side of things. I go back to when I was an operator. I lived within my four walls. I didn’t even have social media. My world was within the four walls. My greatest regret was I wish I would have known more people outside of my little bubble, learned from them. You are one of the guys that I’ve always looked at and thought, I can’t think of a better ambassador for going out and expanding your horizons and the knowledge base that you can tap. Why is that? Does that come from being an engineer, or coming from outside the restaurant industry? Because it certainly seems like you have built real relationships outside of Minneapolis and outside of your restaurants. What’s that done for you?

Shaz: You just touched on arguably the most important lever that I think I have in the capacity of business, and that’s the ability to connect. And per our mutual friend Shawn Walchef, tell your story. There’s nothing more powerful, AI or not, nuclear winter or not, than being able to connect with another human being, because it’s from there that trust is built. And trust is probably the most important thing. Do I trust that this person is telling me the truth? Do I trust they are who they say they are? And is there enough trust for us to collaborate with each other? When the answer is yes to these things, or leaning in the direction of yes, those are the types of relationships that transcend all platforms and technologies. Because at the end of the day, we feel seen, we feel understood by somebody. And I think at the core of it, that’s what we want most, the human need to be connected and understood.

Matt Wampler: That’s easy to say, but how do you break out of the day-to-day crisis of the business and make time for that and find those connections and foster them?

Shaz: Fair question. I think using the tools at our disposal. It’s like going to the gym. You don’t have to do it. It’d be better if you did. But how does one really get started? They have to commit. They have to have the idea, they have to understand the value or the potential of it, and then they have to decide that whether I’m going to be consistent or not is almost irrelevant. I’d like to be consistent, but I have to move. I have to actually take action. And so if connection is important to me, and I understand that there are things that I don’t understand and people that I can learn from, that’s a mindset. And if I establish that, then I’m going to see the world in a way that’s going to be lit up with opportunities. I view every other human being as an opportunity to learn, because there’s a story I don’t know. And the more people’s stories I connect with, the more experiences I share with them, the more trust I build, the wider the net gets. And you know what? When you actually do need something, you’re helping people get where they want to be because of what you know and how you can help. You do that with enough people, eventually someone’s gonna be there for you when you need it, even if it’s as simple as a conversation.

Matt Wampler: As a first principles system thinker, that’s a lot of talking about people and connections and authenticity and trust. How have you found leading restaurants with that mentality? How do you merge the two? What have you learned that has made you an effective leader in restaurants?

Shaz: I think the balance is, you can say what you mean, but you’re not in control of how it lands. And that landing isn’t even in your control, because it’s entirely contextualized around people’s pre-existing biases, the conversations that they’re familiar with, their capacity to manage their emotions when certain things come up, and those certain things have implications in their lives. So knowing how to have the conversation, as they say, it’s not what you say, it’s how you say it. Building trust with people is an organic part of leadership that I think is necessary. I said this at the beginning, ensuring that there is a unification of my character and my reputation. There’s not a person that you can encounter that’s going to tell you, this is who Shaz really is, and it diverges from who you know Shaz to be. And because that’s consistently true, and because it’s consistently important to me, not for the sake of image crafting, but for the sake of me not having to worry about how I’m approaching different people and different crowds, that just makes my life easier. It’s just so much easier to be yourself. And of course, refine yourself when you know that refinement is needed. But the more you do that, the more trust you build. And then people will trust you to have the conversations that may be even difficult or challenging. I have no problem saying, I don’t know how we’re gonna solve this. I genuinely am scared. I’m vulnerable about that. But you know what? We’re in this together. You haven’t seen me do something that’s benefited me and not you. We’re in this together, and I mean it, not with words, but through your experience. Trust me that we’re gonna get through this together too, and that I will always explain the things that need explaining. That alone buys you what you need in terms of balancing doing the emerging tech forward thing and doing the thing that’s right by others.

Matt Wampler: It’s so funny, I tend to agree with everything you’re saying. I tend to be of the mind, and I don’t think this is calculated so much as physically I can’t be another person. I’m not as good at changing, so I am very authentic. It’s worked out for me. But it seems to be really hard for people to do that. It’s really hard to be authentic. It’s really hard to not lie or exaggerate or be vulnerable. Why do you think that is?

Shaz: I think there’s a number of reasons. It can go back to somebody’s childhood. What’s really interesting is that we’re a world full of many adults interacting with each other in high stakes capacities, these billion dollar valuations and superstar CEOs and staff of a hundred thousand people. And at the end of the day, the quality and the efficacy of that leadership can literally sometimes boil down to what was the childhood experience of that particular person in charge. I mention that not to say that’s the sole reason, but one of several contributing factors in the spectrum of human experience that gives people pause or hesitation when it comes to being vulnerable. Even cultural distinctions. There are aspects in society where people will view vulnerability from a particular person or group as weakness, or in a particular context as weakness. And I think it’s up to us, as people who understand what it means to lead people and do great things in the world, to demonstrate that full spectrum. Hey, we don’t know, but it’s okay. Or, I’m scared too. That’s heartwarming, it’s touching, and it feels much more real than anybody who comes my way and tells me, I’ve got this, don’t worry about this. The problem isn’t that they’ve got this. The problem is when they always have that, and I need not look behind the curtain to see what’s actually happening, and then suddenly we’re all surprised. That peeling back of the curtain is the essence of strong leadership. It doesn’t mean talk about everything in every space. It means let people understand that what you’re saying is who you are, is what you’re doing. It’s not a weakness, it’s actually a strength because of its differentiation.

Matt Wampler: Love that. I want to zoom back out a little bit. As somebody that I think is both an insider and an outsider of the industry in a lot of ways, you have a unique perspective. How do you feel about the industry today?

Shaz: I think restaurants are spaces where people connect. At the end of the day, that’s what it’s about. It’s about people connecting with each other. They’re coming for an experience. Yeah, there’s places where you can just get your number three and you’re out the door, and there’s a market for that. But largely, whether you’re fine dining, full service, quick service, brands are built around creating emotions. And those emotions have to mean something. They have to connect with a person’s identity. And that identity never manifests itself better than in light of other human beings. We’re not gonna light up talking to a machine or sitting in front of a computer the way we would with an interesting person or a person that we love or a person that we haven’t seen in a really long time. So for that reason, I think restaurants have a very important place in our society and serve a very particular purpose. And it’s one that we need to defend. I think the industry’s in a great place, because no matter what happens, it could be fully automated Terminator, here’s your slop, but no matter what, people are going to drive towards that human need to get together and share a meal. And so long as that need exists, solutions will be born, whether they be here’s a space where you can do that, or here’s some amazing food.

Matt Wampler: I can’t imagine a more depressing experience in some ways than the greatest plate of food I’ve ever tasted, served in a random cafeteria by a robot, that I ate by myself.

Shaz: Right. That’s what I’m saying. There’s something about it that I haven’t been able to articulate just yet that takes the soul out of it. You know what it is? We’re social creatures. So who am I gonna tell? How will I ever share this in a manner that’s going to create a conversation or connection? I can’t.

Matt Wampler: You want to share an experience with someone and not just your chatbot.

Shaz: Exactly. Right.

Matt Wampler: Well, Shaz, what else gives you hope about the industry? Let’s get back to the AI a little bit. Pontificate on the next ten years. Do you think that this is the beginning of a really rough road where we disintermediate hospitality from food and just become places that serve food? Or do you think this is the catalyst for us to double down on the people side of the business? Or do you think it’s gonna go both ways and we’re gonna see who wins?

Shaz: I think it’s gonna go both ways and we’re gonna see who wins. There’s always gonna be a variety of action taken when there’s a fork in the road, pun intended. There’s gonna be some that say, let’s lean into this, let’s maximize our profit. There’s gonna be some that say, we have to do that, but we also have to take care of people and their experience. And there’s gonna be some that say, to hell with technology, to hell with all of this stuff that’s constantly upsetting every form of my business, let’s just get back to the basics. And somewhere between the latter two, I think, are gonna be the winners. People who understand that people make the industry move forward. Let’s rewind 10 years real quick. The industry may be accelerating, kind of like Moore’s Law of how quickly things are moving. But it was only 10 years ago that we got cloud point of sale. And between then and now, has the industry changed drastically? Sure, the landscape has changed, but at the end of the day, people are still enjoying those meals. People are still making the decisions that they made then, and that’s, do I dine in or do I deliver? It’s still a fundamental decision. And those decisions are still gonna be made 10 years from now, whether it’s a drone or not.

Matt Wampler: I worry about the fact that a lot of our restaurant tech providers are still serving thirty year old software, and AI is now supercharging technological change in a way that you couldn’t have fathomed four years ago. Do you think that eventually it’s Google and Anthropic and all of them that are gonna be providing our technology? You’re gonna get on Claude and it’s gonna be like, do you want to have a back of house system? Great, click this button, we’re gonna ask you some questions and then we’ll integrate it. They’re gonna consolidate and just be the software. They all call it the SaaSpocalypse, the end of SaaS as it is. Do you think there’s merit there?

Shaz: There’s always gonna be merit to what a machine can do in terms of other machines. If you focused all of the intensity towards being able to build a better product, and you know what better looks like and what it doesn’t look like, and you iterate enough times, it’s gonna do that. But I think what doesn’t get accounted for is human variability. Everyone wants something a little bit different. And I just don’t see the overall ethos of many of the providing companies to be, let’s cater this to each and every person in a manner that suits them. It’s more, let’s build something that other people can use to build. Even these models that we use, they’re templatized to a certain degree. There’s a reason they don’t come out the gate with every single thing that we need. Now maybe there will come a day where that happens, but it still feels like customization. In terms of the SaaSpocalypse, I’m actually looking forward to that. And I say that with the utmost empathy for companies that are offering subscription services, but I think it levels everyone up. Those who are providing real value, not just some GPT wrapped feature, not just one thing, not just a set of analytics, are going to actually win here, because they’re gonna understand people and they’re gonna speak to those people about what it is that they want. Those who listen. I think listening is gonna be the key. And those who just build to build, or build to bill, I’m full of puns today, man.

Matt Wampler: I mean, you’re on fire on the puns.

Shaz: That’s gonna be my new gig. But those who just do that, I think they’re gonna go by the wayside, and as they should, to be frank. Because getting back to authenticity, I want everyone in the entire spectrum of my business relationships to be an authentic representation of something that mirrors my values in some sense. Right now there are providers where I may not like the way that they approach the world or their business. I don’t have to like it, but at the end of the day, they provide a specific solution or a specific price that works out for me. I think the day is coming where that choice will be made based off of whether or not our values agree, because I can do something else instead.

Matt Wampler: If innovation is the result of creative destruction and we’re about to get the greatest wave of innovation there is, expect a fair amount of destruction along the way.

Shaz: Well put.

Matt Wampler: Well, Shaz, this has been fantastic. What’s next for you? If anybody wants to come visit you, where do they find you? Where do they find out more information? And where’s the brand going to be in 10 years?

Shaz: I love that question. I’m hoping to be more of a regional brand. We have a ten year target.

Matt Wampler: Global warming’s gonna hurt that. Everything keeps getting warmer. You’re gonna have to keep moving north to keep that cold.

Shaz: Well, we’re in Minnesota, so exactly. Come find us when the data centers release too much heat. We’re still gonna be here, hopefully still headquartered in Minnesota, still serving the most authentic brick oven and New York style pizza that you’ll have, really putting Minnesota on the map, and looking forward to continuing to create spaces that connect people.

Matt Wampler: What makes your cheesesteak the best? Give me the setup when I show up.

Shaz: Yeah, absolutely. So first of all, a lot of people here in Minnesota, we introduce the cheesesteak, the actual cheesesteak. We supply our rolls from Amoroso’s bakery.

Matt Wampler: I’m out of Philly. Come on. All right, you’re off to a good start.

Shaz: There you go. See, you already know. That’s the reaction. We never freeze our meat, it’s always fresh. You gotta cut the onions the right way. We offer whiz too. You can come in and be like, I want a whiz with, and somebody’ll take care of you. We won’t berate you. It’s a balance of offering something that a newcomer would be curious enough to try, and they can switch it up to something that’s more friendly to maybe the Minnesota palate. But once they get comfortable, they can go to the original.

Matt Wampler: All right, I’m gonna get a lot of people in Philadelphia mad at me on this one. But wait, there’s a way to cut the onions. What’s the right way to cut the onions?

Shaz: Well, there’s diced or sliced. And when you’re dicing, you’re doing something that I think is much more in accord with the authentic cheesesteak. Sliced onions, that’s hoagie territory. That’s something else.

Matt Wampler: Okay, well, I’m clearly doing it wrong. I’m a sliced onions guy. I’m gonna have to dice them next time I’m making cheesesteaks.

Shaz: You know the best part about this is you can do whatever you want.

Matt Wampler: Not in Philly. I don’t think they’re open minded for creativity.

Shaz: None really. I’ve seen some of the light poles get greased up when there’s a game.

Matt Wampler: That’s authentic to Philly though. Well, Shaz, I really appreciate you coming on. Again, website, brand, where do they find you?

Shaz: tonomn.com, frankandandrea.com. We have a number of domains, but the point is, you can visit us on LinkedIn, Facebook, Instagram. Actually, visit us on Instagram or TikTok. You’ll see the videos, you’ll see the people, you’ll see how happy people are eating the food, making the food. We’re loving what we’re doing, so come see us.

Matt Wampler: Shaz, you the man. Thanks for coming on.

Shaz: Thanks. Take care.