How Berry AI Uses Computer Vision to Boost Restaurant Efficiency

Jeremy Julian: Today's episode is brought to you by Restaurant Technologies Incorporated. They are the leading automation for cooking oil system out there in the market. If you're not familiar with them, they've been on the show a few times to share some of the unique and cool things they continue to do with oil management. I recently had their newly appointed CEO on the show, and they built an end-to-end closed loop automated oil management system that delivers, monitors, filters, collects.

Jeremy Julian - Restaurant ...: In today's episode, we're talking about drive-throughs. They're a huge engine for QSR profitability. seven seconds you shave off your speed of service is one extra percent of revenue ⁓ you can gain.

The most operators have been managing this for the last 30 years has some huge flaws. If you run a drive-through and you already know your speed of service, Welcome back to the Restaurant Technology Guys podcast. I thank everyone out there for joining us. As I say every single episode, feels like thank you guys, because I you guys have got lots of choices out on the internet to go spend your time and energy.

So thanks for hanging with us. Today, I am joined by a founder and a very, very bright guy who I got the chance to hear last fall. And I'm excited to have Eric jump on the. on the episode.

Really, we're going to talk quite a bit about where things are going and some of the old ways that people used to solve the problem that he and his team have been solving for and then kind of where it's going. But Eric, before we jump into kind of what you're doing with Barry, who is Eric? Give me a little bit of background. Where did Eric come from?

And then we can talk a little bit about what you guys have been solving for because I'm pretty excited to dig in. You know that your team is judged on it. You know that your bonuses are tied to it. And you may just not know how easy it is to fool the existing sensors that are out there.

learned from our guest today that some members wave metal trays to trick the loop timers into reading faster. Cars will pulled forward so that there's no tracking. And it looks like that you've hit your speed of service metrics. ⁓ Two measurement points a customer journey that Jeremy Julian: And recycles your cooking oil.

sounds like they do a ton, but it really is a cool system that I think everybody should be considering looking at. They clean up one of the dirtiest and dangerous jobs in your kitchen. So if you haven't heard of them, I would encourage you to check them out. rti-inc.

com, or you can call them at 888-779-5314. They've hoped helped over. Eric Lam: Great, well, thanks for having me on, Jeremy. My name is Eric, CEO of Barry AI.

I come from a really unique background where it is a family business that manufactures POS devices. And so I did not have a normal childhood like my peers did. I grew up playing around with POS terminals, taking apart and putting them back together. So that from an early age how...

Jeremy Julian - Restaurant ...: actually have six or seven data points that you can now look at with today's technology. In today's conversation, we get into the three flaws that loop timers have and how every operator should understand, including the tricks that staff members will play in order to game the system. We will walk through when pull forward is right and when it is just something that they're trying to play with the metrics.

We also talk about a recent customer announcement Jeremy Julian: 50,000 customers of the United States with their cooking oil. So clearly they are doing something right. And now back to the episode. Eric Lam: I got involved and fascinated into the world of restaurant technology.

And really that kind of paved the way for the rest of my career after I went to business school, returned to family business and had the chance to kind of invent a new chapter for the business. And that's what became Berry AI. So I'm an engineer by school training and then a entrepreneur at heart. Jeremy Julian - Restaurant ...

: that will tell you a lot about where this category is going. A brand you absolutely know and is huge in the industry that you guys are wanna hear about. If you're new to the channel, hit the subscribe button so you don't miss any of the conversations that we have with great operators as well as technologists around the space. Eric Lam: It's very exciting to me to be chatting about the future of QSR here.

Jeremy Julian - Restaurant ...: Today I am joined by Eric Lam, the founder of Barry AI. Funny enough, Eric, like myself, grew up in his family's POS business, making terminals, taking them apart, putting them back together really early, early ages. It was kind of funny to sit and wrap with him.

He also went to Harvard Business School and came back to the family business and built Barry AI to help solve all of these challenges that QSRs have been dealing with with drive-through timing forever. I love it. And you didn't go to just some basic business school. Come on, our listeners a little bit of a heads up that you went to one of the ⁓ business schools in America, I think.

Eric Lam: you I had a chance to go to Harvard Business School. Great two years there, learned a lot. they teach you a lot of things there. They also don't teach you a lot of things there.

So that's the biggest takeaway that I learned. The many things you cannot learn in business school. Jeremy Julian - Restaurant ...: Well, it's the entrepreneurial journey, right?

The figure out ability. tell my, I just had a kid graduate undergrad this weekend and I was like, dude, half of it's about how to figure it out as much as is what they taught you in those books and those lecture halls, right? Let's jump into the episode so you can learn a little bit more. Eric Lam: Yep, that's exactly right.

Jeremy Julian - Restaurant ...: I love it. So talk to me a little bit about Barry. I give us an overview?

What is Barry AI? And again, we're what? Two minutes into the episode, we've already used the word AI. we've got our quote.

We started our quota off for the day. But what is Barry AI at a macro level before we start to dig into the technology and why you think it's so unique and different? Eric Lam: Yeah, that's a great question because AI is the buzzword these days. So it's really important to really dig deep.

What is the AI? What are we doing? From a high level, Barry AI is the vision AI solution provider for QSR. So we specialize in computer vision, which is analyzing what cameras see through ⁓ videos, detecting people, detecting behaviors, detecting vehicles.

In the QSR world, there's a lot of things that we can do with these vision analytics. A example is we would be able to track speed of service really well drive service for a QSR restaurant. So macro level, think about Barry AI as trying to use cameras to obtain more analytics and intelligence from the way restaurants ⁓ operating today. Jeremy Julian - Restaurant ...

: So, Eric, real quick, just for those that are not QSR experts or have never really understood the value of a drive-through, ⁓ I've got somebody on my team that used to work at HME, and he and I will talk about kind of the value of what he was selling headsets for that. And we kind of talked about some of the numbers in drive-through and... for those listeners that either have a drive-through but have never really dug into the analytics of how much more valuable speed of service timing is and all of those kinds of things, as well as just in general, the exponential growth you can have with a successful drive-through.

⁓ talk to our listeners a little bit about kind of why that's such a value proposition to ensure that you're ⁓ the volume that you need to ⁓ analytics and really just in general, how you can increase your top line sales ⁓ you do it properly. Eric Lam: Yeah, I think there's two angles to look at this. So first is to look at it from the restaurant angle. So, you know, quick service restaurants is what QSR stands for.

It literally has the word quick in it. people go to these brands, these establishments with an expectation that they can get in and get out and get their food really quick. So there's a lot of well-known studies into how long waiting impact your sales. ⁓ you know, the, number in the industry is every seven seconds that you can improve in your spirit of service equates to one extra percent revenue that you can generate for your restaurant.

So there's, there's, yeah, that's, mean, people have an actual translation and my very brand to brand, but every brand, their ops leaders will have some way of translating. If we can just shave a few extra seconds off of our drive through. Jeremy Julian - Restaurant ...: That's insane.

Eric Lam: this is how much more revenue we can generate. And the other angle, think, is probably even more intuitive to listeners who may not be as familiar with the QSR space, is just to think about it from a consumer angle. We are all consumers in the market. ⁓ live in an era where we are not accustomed.

We've been spoiled to not wait for things. ⁓ YouTube can a YouTube premium plan because people don't want to spend 15 seconds watching an ad. Jeremy Julian - Restaurant ...: Mm-hmm.

Eric Lam: And so if people are accustomed to this, everything's on demand. can get anything you want within seconds. Just think about your own personal journey. If you've ever been stuck in a drive-through and the difference between a four minute wait and a five minute wait might mean next time you're hungry, you don't go to that establishment.

And so I think that's the easiest, most intuitive way to think about it. Hey, if any restaurant can just convince people that they're moving along a little faster, ⁓ it. it convinces consumers to come back at a higher frequency. Jeremy Julian - Restaurant ...

: Yeah. Well, and, and this isn't, mean, your guys' technology is not, it's new the way you guys are solving the problem, but drive-throughs have been measured forever. We've all been to a drive-through, again, I'm older than you for sure, but I'm older that I remember the drive-through window with the stupid clock and, these guys are rushing around. And so it's not new what it is that you guys are doing from a drive-through timing perspective.

But I guess talk to me a little bit about traditionally how, how people historically had done it. Eric Lam: Thank you. Jeremy Julian - Restaurant ...: loop timers and some of those kind of things.

And why do we think that computer vision AI is such a huge leap forward? I got the privilege to listen to you and the team talk, you know, last fall and I was blown away. I literally walked out of there and I got so educated. when you guys reached out to be on the show, was like, dude, I need the world to hear this because I was so enamored with the different things that you guys consult for that the traditional systems hadn't or can't solve for because of the way the technology was built.

know, 20, 30 years ago when it all got started. So I'd love for you to talk through what is the traditional method and why do we think the computer vision AI is such a leap forward things. Eric Lam: Absolutely. So the way things have always been done QSRs use a technology called loop timers.

They are magnetic inductive loops. Imagine just a metal wire that is buried under the pavement of the drive-through. And so anytime you're in a QSR drive-through, if you're at the pickup window, you look down, you might see a rectangle on the pavement, and that's a loop timer. It basically is a metal detector and it detects when cars are stopped over it.

This technology is quite mature. It's been around for a long time. And it's basically the way that QSRs have measured drive-throughs for the last 30 years. Now, loop timers have always had a few critical flaws.

Number one, they only measure two points in the drive-through because of the way they're designed and the height. kind of labor costs, installation costs, to dig up the concrete and put the loop timers. You can really only put it at the pickup window and at the menu board. And so that's only measuring part ⁓ the customer journey.

So that's the first challenge. The second challenge, and one that we ran into with a lot of our customers, is that loop timers are very easy to manipulate. What's most fascinating is that when these brands, QSRs, try to actually improve some of the service, many times they'll start setting goals or bonuses for their stores that if they can achieve certain speeds on a loop timer, then they get certain bonuses. That actually incentivizes the behavior to manipulate these sensors because they're not really intelligent sensors.

They're really metal sensors. So we've heard of instances of staff taking metal trays and waving it the window and above the loop timer to trigger faster reads on the car. Or another frequent issue that happens is oftentimes the food will not be ready when the customer arrives at the window. And they'll ask the car to drive forward and wait at a parking spot.

And we call that pull forwards. ⁓ Sometimes that's a good operational decision. Jeremy Julian - Restaurant ...: Mm-hmm.

Eric Lam: Many times it's a way of manipulating the speed of service as well to get your car off the loop timer. Finally, the biggest challenge with loop timers as well is that it's very difficult to maintain and it's very difficult to install. sometimes snow or increment ⁓ weather will the loop timers. You need to shut down your drive-through.

You need to dig up the loops and then reinstall new loops. Those are kind of the high level challenges of loop timers. As you expect, computer vision or cameras in the drive-through naturally solve each one of those problems. So when we put in cameras in drive-through, we're able to stitch together the view across multiple cameras so that we know ⁓ the a car arrives in the parking lot, joins the drive-through queue, the timer starts.

We can track that all the way through the ordering process, through the pickup window process. Even if they get pulled forward, we can keep tracking that journey. And so it's really for the first time that the industry can measure the end-to-end guest wait time, which customers, many QSRs believe you got to measure what the customer is experiencing to know what you're optimizing for. So that's number one.

Number two. ⁓ Jeremy Julian - Restaurant ...: No, no, no. was just going to say, I was just going to say one of the things that I, that I also heard you guys talk about is just the flexibility with loop timers because you lack flexibility.

You can't go to two lanes. can't do anything else in the drive through other than kind of force the one path. And so if there's something going on in the parking lot, you get stuck without the speed of service data. So sorry, I didn't mean to cut you off, but it's like, there's a lack of flexibility because it's extremely expensive.

It's a point to point and there's not a whole lot of, you know, there's not a whole lot of variability. So if they're doing something. Eric Lam: Yep. Exactly.

Jeremy Julian - Restaurant ...: in the drive through, you know, they're changing out the oil and the trucks got to be parked kind of somewhere where they might be triggering timing, you're going to miss some of those people if you've got the line that's too long before it you know, you lose out on some of those capabilities. And so, you know, until they hit the menu board, so sorry, I'll let you keep going. But but I just I wanted everybody to listen to hear that says is not just those couple of points.

It is also this whole flexibility because we're getting more and more creative with the ways that people are doing drive-throughs because it does drive, get satisfaction. So, sorry, I'll let you keep going. Eric Lam: Exactly. Yeah, and I'll add one more point to that as well, is that drive-through is becoming an increasing part of QSR's business.

⁓ Many brands are now expanding from single-lane to dual-lane drive-throughs, and the drive-through operation is getting more complex. And the biggest challenge is that oftentimes there's no standardized playbook at the store level. stores are kind of tasked with the mission of, you should adapt how you operate the drive-through. Jeremy Julian - Restaurant ...

: Mm-hmm. Eric Lam: based on the situation. So sometimes it's two lanes, sometimes it's one lane. They need to react very dynamically.

And that's one of the challenges loop timers will run into is, pick one or the other one lane or two lanes. can't just switch between the two that the cameras also address. But back to the original point of some of the advantages of computer vision for speed of service, the other big advantage that we talked about is the the ability to actually provide accurate data that cannot be manipulated. So traditionally, if staff were pulling cars ahead, that timer doesn't stop in a computer vision world.

If a staff is waving a metal bin above the loop timer, the camera's not fooled by that. And finally, ⁓ installation and maintenance of the cameras is a lot simpler and easier than loop timers. You don't have to shut down your drive through. Jeremy Julian - Restaurant ...

: Mm-hmm. Eric Lam: If one of the cameras goes down, rest of the ⁓ cameras and the timer still works. overall, I would say it's easier to maintain a more robust system. Jeremy Julian - Restaurant ...

: Well, Eric, you said that there are certain operational reasons why you would want to do pull through. again, I got to hear you talk on stage about why you would want somebody to pull through versus not. ⁓ I guess I'd love ⁓ ⁓ to your thoughts on what makes sense operationally to pull somebody through versus not ⁓ how really computer vision can help ⁓ that transaction to that that got pulled through versus the one that shouldn't have gotten pulled through. So I guess why don't you educate everybody?

⁓ What a good Eric Lam: Thank Jeremy Julian - Restaurant ...: good sense of pulling somebody through. Why would you tell them to go pull into a parking spot or pull through to the front of the drive through so that you can run food out to them? And how can you with computer vision, you know, like really even tell whether that was the right thing to do versus not, whereas you can't with the loop timers.

Eric Lam: Yeah, so the answer is that it depends on the concept. And I can kind of give a few examples. So generally from a high level, you want to pull the car forward when that car is going to have a long wait time. Now, depending on the concept, that long wait time might be triggered by different things.

For example, if they're ordering a very large basket, very large meal, the equivalent of five or six people's worth of meals. in order to keep the line moving, that specific item, chicken fingers, fries might not be ready right away. so in scenarios like this, it's often best practice by these brands to pull the car forward so that they can actually start serving the cars behind that big order cards. For some other customers, many brands today are cooked to order.

So they kind of promote that they are the most fresh, food quality. They don't prep things ahead of time. ⁓ in instances like that, when ⁓ cars you are actually pulling most of the cars forward because the food is being prepped only after you order. And so in scenarios like this, you want to make sure that your staff are incentivized and they're trained that they're also pulling these cars forward.

Now, where you want to draw a fine line is when there's not scenarios like this and it's just someone, a staff kind of manipulating the speed of service times. And the way we tackle that from a computer vision standpoint is that we're actually able to look at both the POS data as well as the drive-through information in real time to know, hey, was this car being pulled because there was long line behind them and the staff probably wanted to get the line moving. Or was the car being pulled when there was actually no car?

There was no car in the drive-thru. It's PM at night, and there's really no excuse for why this car should be waiting. There's no car waiting behind them. And so we're able to differentiate between those and also look at the POS data and the receipt data to these brands look at, ⁓ these forwards kind of by the book, or were they more likely to be something we call like a flag pull forward that's potentially a manipulation of the timers.

Jeremy Julian - Restaurant ...: Yeah. Eric, one last thing on this whole kind of thread before I jump into some of the things that people say from a downside on it data amazing. ⁓ Data amazing ⁓ if can do something with it.

But the thing that I continue to talk to ⁓ operators and that are building technology is ⁓ how we make this data actionable? ⁓ How we make it so that we can solve true guest problems as well as restaurant problems with this? because just throwing a whole bunch of data at it and changing it from loop timers to this, they're humans, they're either gonna find a way to go manipulate these systems and or they're gonna be inundated with data without the ability to execute against it.

So I guess I'd love ⁓ for you to talk because again, I heard you ⁓ and team tell some stories ⁓ last fall that I was blown away with that it's not data for sake, but it's truly helping make the business better. ⁓ both from a guest perspective as well as from a staff perspective so that it's not this like, ⁓ know, they just want, want, want, want what, you know, they need to be able to solve that problem. So I'd love for you to share a little bit about how you guys are using the data that you have to create a better guest experience as well as a staff experience.

Eric Lam: Absolutely. My biggest take, and this is a big learning for myself as well, ⁓ I've been in this industry, is that less is more. And is really something that took years to, I think, internalize what that meant. And I it manifests in two ways.

So I think when we think about how do we make this simpler and more actionable for restaurant staff, the first approach we take ⁓ is Jeremy Julian - Restaurant ...: Mm-hmm. Eric Lam: We want to make sure we can deliver information in real time because when you deliver in real time, then there's actually enough time to fix it right away. and so, ⁓ know, part of our product is a real time dashboard in the restaurant itself.

And iterate on the design a lot. We want to make sure even for someone who doesn't fully doesn't have, for example, great command of English, they can understand this dashboard. should be so simple. sort of like your IKEA furniture assembly instruction guide, that someone with no training can look at your dashboard and say, ⁓ you know what?

I can see eight cars in the drive-through. So I don't need training. I know that means I'm going to run out of fries in five minutes. And so it took us a lot of iteration in terms of what are the best metrics?

What are the easiest ways to present this information so that it doesn't take much training for staff? in poor and QSR when there's such high turnover. So the first angle we looked at things. The second is from the detailed reporting that restaurants get from our system at the end of every day and at end of every week.

And that's really where, ⁓ you initially ⁓ without careful, we inundated our customers with too much data. We gave them every single metric they could have possibly ever wanted for. And it took a deep collaboration with our customers to start cutting down from that and saying, Hey, we only need to focus on these three or four metrics. And for all the rest, that's great.

Let's put that somewhere else. If someone's really passionate about it, they can go look at it. But for the most standard reporting, let's really boil it down to, customize metrics. And, know, at Barry, we really take it even one step further to reduce that training burden.

Everyone different terminology. for the same thing. Some people call this experience time. Some people call it total journey time.

Some people call it customer time, et cetera. So we also make sure that in our system, it's designed to be flexible enough so that we can customize it to the customer's or their terminology that they're very used to. ⁓ all these little things are what we pay attention to so that when people look at our system, it's very where the bottleneck is. Jeremy Julian - Restaurant ...

: Yeah. Eric Lam: We like to say that our system, our goal is not to get you to spend more time on our dashboard. Our goal is to make it so that you can get a quick glance and know what you need right away. We're not optimizing for how much time you need to spend on our website.

Jeremy Julian - Restaurant ...: I love that. Eric, I'm going to take a pivot real quick and ask about, you hear a lot about computer vision and the part that I guess I hear from consumers or customers is how do you deal with the privacy factor? How do you deal with the fact that you've now got cameras?

The loop timers was fine because you knew it was just measuring the bottom of the car. You've got certain states, certain legislation that says, we can't capture images and certain you know, things. for those listeners out there going, it would never work in my state, in my store because of that. I'd love for you to kind of, guess, dispel that myth that says we're not capturing PII.

We're not doing any of that, that kind of stuff. So love for you to talk through a little bit of that, cause I'm sure you get that, get that question from time to time. Eric Lam: Yeah, that's great question. I would have a few advice for anyone out there who's thinking about this privacy question.

first is really be very specific and be ⁓ narrow in terms of what the use case is. What are you using cameras for? It's not that ⁓ form of camera recognition is treated as same. ⁓ ⁓ personal identifying information, is really the key here.

camera identify any identifying information, whether that's ⁓ who are, like facial recognition or vehicle license plate ⁓ on car, those are all kinds of identifying information. So that's not illegal. just ⁓ you have more disclosure if you're doing that kind of use case. where we've kind of been very intentional about spending our time is making sure we don't kind of take that step into identifying information because quite frankly, there's so much value that we can do even just from an anonymized fashion.

So we need to know, hey, this car is Eric, this car is Jeremy, or this car is license plate ABCD. We just need to know this red car ⁓ spent minutes and 30 seconds in the drive-through. And so when it's use cases like this that are anonymized, not identifying. Your cameras, at least as of today, are regulated in the same way that security cameras are.

The other piece of advice I would have is obviously, AI is evolving quickly. Different states are having different approaches to it. As of today, no state is going to say, you are Just because you have cameras means you're capturing personal identifying information. And so my other advice is if you're engaging with a camera vendor or a computer vision vendor, it's worth asking them about what's their understanding of the latest regulation this because it is changing.

The last thing mention that ⁓ we take an step of precaution at Barry AI is that we actually don't keep any of the videos. all of the videos that we process from the cameras, they're processed on the server inside the restaurant. And once the AI analyzes all the cars, all the journeys, that video is no longer stored. And so from a regulation or compliance standpoint, that's actually easiest for ease of mind for restaurants.

Jeremy Julian - Restaurant ...: Yeah, no, and I love that. Eric, I know we're recording this the first week of May in 2026. I knew that we were waiting till earlier this week to have this recording, partially because you guys had some pretty exciting news.

Funny enough, I got the email from one of the QSR magazine, I think, ⁓ hit my inbox, and I knew that we were getting started to record. I guess I'd love for you to share a little bit about your guys' huge win, because ⁓ it's a pretty big one. It's a pretty big one and I guess I'd love, you know, it's already kind of in the press release, but for the listeners that may have missed this, who did you guys just land and kind of what drove that decision for their brand forward with Barry?

Eric Lam: Yeah, so the big announcement that we had a few days ago was that Culver's restaurants has announced that they're going to deploy Berry Eye to their stores nationwide year. And so that's over a thousand locations and it's another brand in the QSR space we have chosen to lean kind of completely into the computer vision technology ⁓ and use this as opportunity to leapfrog some of their peers. The Culver's journey, know, we're very grateful to be working with them. They're very forward thinking leaders ⁓ the team.

They're actually coming from a space where previously they were not using loop timers. They were not using these of systems. So to them, they did not have the visibility. to really think about and measure speed of service.

has always been a restaurant that prides itself ⁓ ⁓ cooked fresh, made to order. But it's expanding kind of nationwide and expanding into new markets, it's entering markets where customers may not be as familiar with the Culver's brand. And they don't know that it's cooked to order. They don't know that.

When you go to Culver's, you need to wait, you know, five minutes, 10 minutes for your burger, even though it's delicious, it's a bit of a longer wait than some of the other brands. And so we chatted with the Culver's team, that was some of the thinking behind, ⁓ why invest in new technology like this? Why really focus on speed of service is because, you know, in new markets, when no one understands your brand, that's basic expectation that folks have. And so it's a, it's a It's a long journey that we worked with Culver's.

It started with small scale pilots, and then we expanded to larger pilots. And throughout that journey, incorporated feedback from pilot franchisees and tweaked the product, adjusted our terminology, et cetera, and really made sure to find a way to make it work for their needs. So that was a big announcement a few days ago. Jeremy Julian - Restaurant ...

: Well, congratulations. And ⁓ I've got a butter burger on the way. I got to go figure out how to go find me one. There's a call where it's maybe 20 minutes from me.

I got to head up that way later this week. So have to check it out. Eric Lam: There we go. Yeah, my favorite's the mushroom Swiss.

Jeremy Julian - Restaurant ...: Yes, that is a pretty darn good burger. A ⁓ Culver's meal for me is not complete without some cheese curds, so there's that. love it.

Eric, talked a lot about DriveThru. guess, where is the ⁓ computer technology going? We've had a guest on the show ⁓ who really doing inside the store stuff. ⁓ Are guys trying to stay outside the store on DriveThru?

Are you guys trying to work on some of the speed of service ⁓ and of the other areas that you guys can improve the guest experience? Eric Lam: Yeah. Jeremy Julian - Restaurant ...: within the four walls or even in the drive-through, guess I'd love to listen to you kind of riff a little bit about kind of the product roadmap and where do you see not only this technology but very AI going to help you really create better guest experience, better staff experience for restaurants out there.

Eric Lam: Absolutely. So we kind of described three chapters to the Berry Eye journey. And we really just finished chapter one and now entering chapter two. So when we started off, we really wanted to find one specific use case that cameras could solve and that restaurants would care about and really adopt.

And that to us was the drive-through timer. And so We've been very focused on that. We're very grateful to be working with customers, Culver's, Zaxby's, other brands who have fully adopted this type of drive-through camera timer. Chapter two is really where a lot of our customers are now kind of guiding us or pulling us towards is they say, hey, you know what?

I don't want two camera systems in my restaurants. I have one set for the drive-through, for the AI, and then I have another set of security cameras inside the restaurant. They're two camera systems. They don't talk to each other.

And so, you know, Barry, can you guys figure something out and really unify this experience? so chapter two for us is really becoming that security camera provider as well inside the store for the restaurant so that these restaurants can have a unified experience with their cameras. if they see a very slow speed of service, a certain day part, then they can pull up the video right away of what was happening in the kitchen. or they can pull up the video right away of what certain item was not prepped ahead of time, et cetera.

So that's really chapter two of our journey. And what excites us the most, I think, is chapter three, which is when you have a lot of cameras, there's really a lot of things that cameras can be detecting and analyzing. cool about this is that at this point, the restaurants really become your thought partner because they'll ⁓ Jeremy Julian - Restaurant ...: Mm-hmm.

Eric Lam: very inspired, hey, can you also track how many boxes of inventory that were just drop shipped? Hey, can you also check how long the fries were sitting out? Can you check if the tables are clean, if we're taking the trash out on time? a highest level, the way to think about it is, from brand standards, operation standards perspective, there's hundreds of tasks that you need to be checking manually and auditing.

Any of those tasks that you can do with a human eye, is something we eventually want to be able to help with with cameras because one signs up to do this job to be auditing how long the fries have been sitting, you know, in the holding bid. Those are things that we believe cameras are really well suited for. They can work in the background. They can work quietly.

They can audit things based on defined SOPs and that will allow brands to let their staff go focus on what brings better experience to the consumer. and actually spend their time interacting with guests and taking care of the employees and not having to check how many boxes were dropped off. So that's high level how we think about how we evolve as a company. think really, we're just getting started with what cameras can be helping folks do operationally and inside of restaurants.

⁓ Jeremy Julian - Restaurant ...: Yeah, no, and I think it's, ⁓ I think for me, it's one of those amazing things that the cameras have gotten inexpensive enough. They've gotten, you know, easy enough to install and the data is so good and so rich. And it's not essentially, I've heard people go, ⁓ you're just going to slap people on the wrist.

But at the end of the day, you're getting them to get to the right place to be able to do the job that they need to do and not. get stuck, you know, managing checklists and doing those kinds of things. So it's automating so much of that so that they can truly be in the hospitality industry because that's why we go out to eat is to have an experience to enjoy the food and to, create that guest experience. so if the, if the computers can help train them on how to do those things better, I think all of us will be better off with the team members in the store as well as the guest experience.

So I love that you guys are into that Eric. Eric Lam: Yeah, absolutely. Jeremy Julian - Restaurant ...: Awesome.

So talk to us a little bit. How do people get in touch? How do people learn more? What can they expect the experience to look like if they choose to, you know, they got to the end of this and they're like, okay, I'm 30 minutes into this.

I need this for my drive-through. I need to figure out how I can get what Barry's doing. Who are your guys's, you know, kind of clients that you guys are looking at? How did they engage and what would the experience look like if they were to reach out?

Eric Lam: Yeah. So open to working with any QSR, whether it's drive-through or you don't want to measure speed of service inside your restaurant, we can also do that. The best way to get in touch with us is to come to our website, Barry-AI.com.

⁓ You'll be to see different case studies, different product ⁓ ⁓ And you in with us, we'll jump on a demo give you of in-depth. understanding of this is what the product will deliver inside the store. This is what your reporting will look like from an email standpoint. This is what you'll be able to see on our web dashboard.

So you can really get a feel of what the product is and what ⁓ ⁓ get out of it. And we'll talk about letting you try it out. Most folks try it out at ⁓ or locations to start and really that's when they can see, what do we get from this? How do we act differently if we have this information?

⁓ That's the best way I encourage folks to get started is to try it out. It's pretty low cost to try out and are there'll be a good way for you to think differently about how you're in your drive-throughs and how you identify bottlenecks. Jeremy Julian - Restaurant ...: Yeah, and I say it all the time.

If you're not doing it, likely your neighbor is to your right or to your left. And so I would encourage anybody that's got a drive through, anybody that's trying to measure speed of service in a different way. Eric has talked a lot about kind of what they did and why Barry exists and how they're really solving these problems. So for listeners guys, we know that you guys have got lots of choices.

Thank you guys for hanging out. Eric, thank you so much for jumping on and to our listeners, make it a great day. Eric Lam: Absolutely. Thank you, Jeremy, for having me.