Revolutionizing Restaurant Operations with AI-Powered Video Analytics

In this episode of the Restaurant Technology Guys podcast, Jeremy Julian sits down with Brock Weeks, CEO of Savvy, to discuss groundbreaking AI video analytics transforming multi-site restaurant operations. Brock explains how Savvy helps brands unlock valuable insights from video data, covering speed of service, guest experiences, staff performance, and delivery efficiency. They delve into the evolution of video systems from liability tools to essential business intelligence assets, emphasizing the importance of integrating video data with other business systems for full operational visibility. By addressing common challenges such as fragmented data and leveraging AI for actionable insights, the episode provides a comprehensive guide for restaurant operators looking to enhance their operations with technology.

00:00 Savi

00:14 Introduction and Guest Introduction

02:18 Brock's Background and Savvy's Mission

03:18 The Evolution of Video Technology

03:40 Challenges in Offline Retail

05:50 The Power of AI and Video Data

09:55 Current State of Video Systems in Restaurants

17:30 Business Use Cases for Video Data

21:24 Understanding Data Centralization

21:50 Identifying and Solving Operational Issues

22:55 Enhancing Guest Experience with AI

24:30 Revolutionizing Store Operations with AI

26:26 The Future of AI in Retail

30:08 Implementing AI Solutions in Your Business

36:11 Connecting with Savvy for AI Solutions

Transcript

Speaker: 0:02

This is the Restaurant Technology Guides podcast, helping you run your restaurant better.

0:14

Hey everyone. Welcome back to the Restaurant Technology Guys podcast. And today's episode is a really cool technology that's a game changer for operators that turn video into data that drives business intelligence. Today I am joined by Brock Weeks, who is the CEO of Savvy. They have some really cool AI powered. Video analytics that's really changing the way operations can happen within a multi, multi-site restaurant group. He shares how savvy is helping brands move beyond just cameras for liability and helping them to unlock speed of service, insights, guest experience, insights, staff performance, insights, even how to make delivery. Flow better. We talk about how AI is not just a buzzword, but it's really pinpointing the root causes of bad reviews and opportunities to coach your team so that they can get better, and you can ultimately have the best guess satisfaction in the industry if you're wondering how to make it work for you. Just sit and watch this episode all the way till the end where he challenges us to look at scalable solutions that are incredibly affordable and likely are going to be something that you're already paying for. If you don't know me, my name is Jeremy Julian, I'm the Chief Revenue Officer of CBS North Star. We wrote the North Star Point of Sale solution for multi-units. Please check us. At CBS northstar.com, if you're not already a subscriber, please do so on your favorite POD podcast platform, including YouTube, Spotify, and Apple Podcasts. Have a great episode.

Jeremy Julian: 1:42

Welcome back to the Restaurant Technology Guys podcast. I thank everyone out there for joining us. As I like to say at the onset of these, you guys have got lots of choices, so thanks for hanging out today. I am joined by an industry colleague that I've met before, but really it's only been like the last couple of weeks that he and I have gotten, gotten to know each other a little bit better. And, I love the timing quite frankly'cause, what Brock's gonna talk about here. As to what they've been doing is really gonna blow people's minds, I think. And I'm excited to, to have him share. But Brock, why don't you introduce yourself to our listeners. Who's Brock? what does he get a chance to do professionally? And, long walks on the beach, romantic novels, all of that kind of stuff. Just to introduce yourself to our listeners out

Brock Weeks: 2:18

Yeah, no, thanks for having me on Jeremy. And we can handle that. outside of the long walks on the beach, I'm fortunate to be the co-founder and CEO of a company called Savvy. Where we're working to unlock the value of video data for multi-site restaurant and retail operators. really just helping them understand what's going on in their business so they can operate more efficiently. And, more important than all of that,'cause what I get to do daily, right? I get to work with a lot of really great people, and more important than the people that I get to work with. This is my family. I'm married to an amazing wife, Kelsey, that we're gonna be celebrating 17 years, here, this, month actually. Yeah, thank you. And, 17 years, tons of memories. Three kids now a nine 11 and 13-year-old that are really like the most important part of my life and get to spend a lot of time with them and love the space that we're in, right? There's nothing better than, the hospitality industry. Not only just'cause of the people, but it's like brands that we get to interact with. Like my kids can be involved with what I do for work and get excited about it. So it's ton of fun.

Jeremy Julian: 3:17

I love that. so Brooke, you I don't say you, you went through pretty quickly, what is video and kind of talk to me a little bit about the advent of video. the company's not been around. it's been around for a while. you started and you guys have been doing things, but I think there's. Today and even really the history of kind of video. There's been a lot of advancements. So talk us through what is video and why Cho? Why choose offline retail? Because when you and I talked a couple of weeks back when we were preparing this, it was like the problems that we have in offline retail are a challenge. Getting data, integrated back on a multi-site. So I'd love for you to walk through what is savvy as far as a product and then we can dig into what problems does it solve that people haven't, haven't, haven't been, been doing in the past.

Brock Weeks: 4:00

Yeah, no, absolutely. I think to start, and my background is when video started to be utilized, in spaces was I was fortunate to join a pretty high growth, technology company that was focused on the home automation space right out of graduating from college. where I gotta spend some time in their sales teams and building sales teams for'em. And then, really informative to, they received some capital investment from Blackstone to be able to go on the strategy side and really look at how technology was changing. And what we saw there, which led to starting Savvy in the future, was the fact that these security systems that had been installed in homes and had a single point purpose. Because of the dividend of the smartphone wars and technology and the processing power, really getting extremely affordable. You started to see these sensors have multi-use case and because of the internet and them being able to talk and communicate to each other, that alarm system now could be used to do things like control your door locks, control your lights, control your thermostats. So you were able to now start to do affordable automation systems with hardware that used to just be used to lock your doors and set an alarm off. Because of the advancements of all of these other technologies, these sensors now became very valuable to a homeowner. And we saw the same thing happening with video, which is what led me to start savvy. Which is, it really came into prevalence in the early in 2000 tens. IP video got really popular around 20 14, 20 15, and people started changing from the old DVR or VCR to tape-based systems and putting in these really robust, solid, recording systems. But it was still typically being used for single purpose, which was liability protection.

Jeremy Julian: 5:49

Yep.

Brock Weeks: 5:50

But as the rest of technology started to advance cloud computing cost and storage cost and AI analysis, which really at the end of the day, AI just allows you to structure large, massive amounts of data and gain in and out of machine, tell you the insights out of that data, right? It structures it for you, so it's more useful. And we saw that. This industry above every other industry was distributed, which meant it was really difficult for teams that weren't physically there. To be able to understand why they were receiving the results they were re achieving, and there was always variables amongst the different stores in the same geographical area. You would talk to the managers and they knew what a good store should look like, but they would get reviews. It became really popular to get, let's interview all of our customers. if I go and have a bad experience, because I'm somewhat lazy, like most people, all I'm gonna do is say service was slow. I am not gonna take the time to say, Hey, you actually had enough people in the restaurant. They just weren't doing the right task. One person was doing this, one person was doing that, and that left me to have this experience and I waited forever. Here, no one does that, and by nature of that, operators are left saying, okay, we know we have a symptom. What's the actual root cause? Crazy difficult and time consuming to do that.

Jeremy Julian: 7:20

and historically what people will do is they'll put people in stores and they then secret shop and they spend all of this time and all of this energy trying to figure out what is the root cause of these items. They'll move managers around, they'll do all of these different things and to your point, they have the data at their fingertips. If they have the right tech, it sounds

Brock Weeks: 7:37

A hundred percent. The vi the actual interactions are all recorded. A lot of'em have audio, so they have the video and audio of exactly what happened. Our premise was though, is the data video was unstructured, meaning the only way you could structure it and search it was based upon time. We looked at it and said, okay, if AI is going to become as, powerful and as accurate, and as affordable as the technology trends are showing. What do you need to do with video so that it can be used by these operators because they have unique needs based on the fact that they're distributed and it was, they need to be able to move. Like any data set, it needs to be moved to the cloud because if it's reliant on customized hardware, and that's kinda the trend in any technology, it starts off very specific, very customized and expensive piece of hardware that does one thing. Over time, it gets better at doing those things and more affordable, which you've seen in the hardware game. you can get a phenomenal camera now for anywhere from 80 to$200 a month that like, or not a month. it, that's it's cost the camera. but if you had that happening and you need to be able to move it to the cloud, and you needed to be able to actually structure it by your business data. Not by time because you don't exactly know when all these things happened. You just got the guest review, but you don't know exactly when their experience happened. But if you have your other business systems and you could tie that review right to the point in time and search your video data by that review, now you can answer those questions really fast. Even if AI doesn't come there, but you're able to integrate it to your other business systems, you can start to use it. and the last piece, was that, it needs to be able to be given in a digestible way. To the team because a store manager's really good and a district manager at running their business, but they're not financial analyst. I'm not a financial analyst. Like you can throw me a spreadsheet and it takes me a while to digest it. So it needed to become in a digestible way that you were actually not removing the people to spend less time in the stores, but helping them be more effective when they were in the stores by giving them here is the root cause. Now you go coach, mentor because you know how to do that better than anyone.

Jeremy Julian: 9:54

I love that. and so I'd love to, to walk through Brock, just what is the state of brands that aren't using a tool like Savvy? you talked about the history and again, I go back to some of the earlier video days where it was a big DVR, it's in Iraq, it. It's, single point. You might be able to dial into it over the internet and go scroll through those things. But it's not cloud enabled, it's not event driven. It's this big box that's there. It's connected to these proprietary cameras that, that, have these proprietary lines. That's historically what I've seen. And then I've got my home camera system, which is these wireless devices that go all over the cloud. Check my kids. And what time did they get home? Their curfew was this time I was already sleeping. They should, your kids are young enough that I got three teenagers at home. Dude, I, I gotta make sure that they're in bed when I get up at five 30 in the morning and make sure that they got in during curfew. So I look at my wireless cameras, but there's a big diversions from what I see oftentimes. Inside the restaurant brands, is that kind of still the state of the majority of people, or are a lot of people that you guys are working with kind of at that state? I'd love for you to talk through prior to Savvy, what are you seeing out there? talk to our operators that are sitting here listening, going, yes, this feels like me, and get catch me up on the state of where things are prior to getting to savvy.

Brock Weeks: 11:05

Yeah. because it was typically just an expense, right? A liability expense, an insurance policy. It was pretty much okay, great, I'm just gonna go get a system that records and that I can pull up and look at. and the reality is, especially from the mid-market on down, right brands, that's the state most of'em are in, and they are trying to solve that one problem. And I would say that.

Jeremy Julian: 11:27

The slip and fall, the underage drinking, the, these things. that's what you said, and I just want to, I wanna speak to that operator that's out there that the only reason they use the camera is not to make their business better, but it's to make sure their insurance policy, if something were to happen, they've got the footage to be able to do that. is that what you're seeing out there?

Brock Weeks: 11:43

A hundred percent. And is that important? Yes, it's very valuable, right? It's one of the things. But what we are seeing now that ai, I think it's pretty, Ubiquitous across anyone's in the industry. It's here. And the power of AI is like it's real. There's a lot of refinement and a lot of learning how to use it, and it's only gonna keep getting better. And at a rate that is just exponentially increasing, but it's here. when any new technology comes about and we don't quite know how to use it, we go to the default, easiest path, which is the liability. then the operators start to say, oh, great. I know I can UI see the vision of how I can use video. And what they instantly do is go to a product that solves an existing problem that they have. So it may be speed of service, right? Either it's an iot device of a loop system, or they put in a really customized camera system to track drive through speed of service or in retail locations, they do guest counts and conversion. And once again, really valuable. But then they start to see all of these other different use cases. And so what we recommend to operators, whether they work with Savvy or anyone else, but it's how we approach the problem is that this data set. Has immense value to your organization and should no longer be neglected. And in order for this data set to provide value, there really needs to be three things it needs. You need to get a really stable system that's gonna run for years, right? You're gonna get your ROI investment. You probably don't want to go buy a$20 system at Best Buy or Sam's Club, that's consumer grade for a home'cause it's gonna be used differently. But really affordably, you can go get just a good, economical, reliable system. It's gonna last us seven to 10 years. Make sure that you're covering the areas that you would have business use. Case around, not liability, right? Business use case. And liability. Come up with a brand standard, figure out exactly how you want to do that because a sensor is a hundred bucks. It's more expensive to have someone go set it up and when they're there, they'll do it really affordably. If you're having to come back every time, you're gonna cost a yourself a lot of money. And because they approach the problem back to your question of what we see to just solve a product, we see them have. Between two to four different types of cameras or sensors with two to four different servers, all storing video data and all processing and analyzing video data for different points of the business with different functionality. All of them need to be managed as a separate vendor, as a separate piece of hardware, as a separate portal, and the data is stored separately and siloed and fragmented.

Jeremy Julian: 14:28

That sounds awful. Sorry. I'm like, you're not getting mu much business value unless you are just the person that's the risk manager or you're just this, or you're just that. If you're having to do that, I'm sorry to cut you off, but it's like that blows me away to think that they would have that, but. It's not uncommon for people to solve, they've got a, they've got a pain on their finger. They're gonna solve the pain in their finger. They got a pain in their foot, they're gonna go to a different, so I love that analogy'cause I didn't realize that was, that it was so disparate.'cause you never are gonna have somebody that's gonna have one POSI mean, we work in the p os space. They're not gonna put a POS in the drive-through, in a different one at the front counter. But it sounds like they're doing similar things. They're gonna have one for their drive through. They're gonna have one for their front counter. They're gonna have one for the back walk-in. that's insane to me. But I'll let you keep going because it's

Brock Weeks: 15:08

No, you're, it's back to the days, right? Of the different iPads for all the different third party deliveries. And it's no one, anyone could just look at that and say, this is not a good idea. Like so many fell points. And if you're looking at it video, because it does answer all of these questions well, even if the AI is as powerful as could be. It is only as powerful as the data set it's applied against. So if it's always only looking at this fragmented

Jeremy Julian: 15:32

Yeah. If it's only a pars parsed piece of data, then it makes it really tough to deal with

Brock Weeks: 15:36

Correct. It's like when you go into a brand and the drive through's humming because they're measuring the speed of service, but you stand in the lobby for 15 minutes going I got an ice cream addiction. Like I just want a frosty folks. let's take

Jeremy Julian: 15:46

Yeah. Yeah, exactly. I'm gonna go get in my car and go get it faster to sit back down in the dining room.

Brock Weeks: 15:52

Yeah, because once again, if you only look at one piece, you don't get the holistic view. So that really is what we're seeing, and it's happening at small brands and it's happening in large brands. And so whether it's with Savvy or anyone, it's. Look at this as a techno, a key piece of your tech stack and a key data point. If what we've talked about, like you're like, yeah, I would love to know those things in my business. Great. The data needs to be aggregated and you typically, you've seen this in point of sale. Where the hardware has essentially become separated from the software. If you looked years ago, you had to buy this whole package where everything worked together. It was very hardware focused. Then cloud computing and all of that allowed the software to separate from the hardware to now most of the point of sale vendors, their hardware is not proprietary. It's right. Then they may be interchanging different hardware and different software companies. I think you need to take the same approach because. The most expensive part of adopting any technology is disrupting the store and having to change it. And if every two years you're having to do like a rip and replace disruption, it's gonna get difficult. So get a great hardware layer, centralize the data set, and look at it as a, a piece of, a true piece of the data stack and as a dataset rather than a product. And then you'll be in a really good position.

Jeremy Julian: 17:12

Yeah. I love that, that thought. So Brock, I'm gonna, twist it around just a little bit. You talked a little bit about where things started. We talked a little bit about the evolution of, some of the hardware stuff and just how much easier it's gotten to deploy. Talk to me a little bit about kind of some of the big business use cases you're seeing outside of using AI to ask the questions that you need. Risk management, where everything started, but you're, you talked about some things that I think our listeners may not have considered that you are seeing. Drive business value to their business as they implement savvy? Is it speed of service? Is it, guest satisfaction? Is it staff member stuff? I, you and I talked offline and I'd love for you to share two or three examples of things that you see so few operators doing, and then when they get it, it just drives monumental results inside of their business.

Brock Weeks: 17:59

Yeah. we'll start with, that guest experience. We chose to take that as one of the first kind of key things we wanted to tackle because as we looked at reviews of restaurants. Three quarters of all negative reviews reference the speed of service, the interaction with the employees. Or cleanliness of the restaurant. Those were like the key pillar cornerstones of most negative reviews. It wasn't actually the quality of the food, it wasn't those things. and so we looked at and said, Hey, can video actually move the lever here to improve the quality of the operator's lives and answer these questions? And the question, the answer was yes. So speed of service, not only, there's a couple different brands where we've got the case studies published on our website. That, before they were spending anywhere from six to$15,000 a store to install a different type of loop system to measure the drive-through speed of service. studies show those are in the high 70% accurate. It's basically vehicle in, vehicle out, and if anything gets off, it resets every 30 minutes so that it can recalibrate. as everyone knows, really expensive, only showed a part of the journey. and so employees, better or worse, when you start to measure things, they start to try to game the system. And we've all been there where you've been held up beginning to the menu board to actually order, because that's when the timer's gonna start. Or you've gone to pay and your order's not ready yet. And they're like, yep, can you just pull forward over to there? And then you get forgot about and left. You have to get outta your car. Like we, we've all had these experiences. And what we're saying is most of these brands have installed cameras out there for the liability protection. AI can now accurately and affordably track the entire guest experience, and you can actually segment it by just drawing regions of interest digitally on the camera and start tracking not only the holistic view of when they enter the lot and when they leave the lot. But also the actual different steps within that process. So you can start to really refine where do I have bottlenecks? Because it's really just manufacturing capacity and throughput capacity. And so when you start to provide your team more insights and visual data by measuring these things, and it's done in a way that it cannot be gained, the focus moves to actual process improvement instead of crisis management historically.

Jeremy Julian: 20:18

Or I'm gonna get yelled at.'cause I moved this car up and then they got pissed and then they gave me a bad review. It's what caused me, what order did they place that caused our kitchen to get backed up? What is happening there? And I love that you said that's, because that's. That is, and again, historically with a time-based system, you'd have to look through hours and hours of tape to figure out what happened there. Whereas now with some of this advent, it's Hey, what, I guess I'd love that query, what is that question that you can ask the AI that says how many orders? I, just share it with us and I promise you, because you and I talked prior, I hit the record button. Some of our listeners are gonna be like, what? You can do that today?

Brock Weeks: 20:53

Yeah, it's once again, measuring time and segmenting it. That's, once again, it's more just giving you the baseline. you can just hit a button and be like, show me the five longest wait times, and then watch the entire process of that.

Jeremy Julian: 21:06

Yeah. And people are gonna be blown away, Hey, what, who's anybody that's been over four minutes in the drive through, six minutes in the drive through, eight minutes in the drive through, and you just look at that, it's now I can see all of the videos and it's Hey, on Tuesday we had a really crappy shift. We on Wednesday we did great. And so I, it's so much simpler than people make it out to be and I appreciate you sharing that.

Brock Weeks: 21:24

And because you've centralized the data of your business systems while you're watching that experience, you have the transaction data. So you know. Was it just a really large order that threw us off? Because that happens, right? And you can't, or was it, Hey, we actually got the order wrong, and so we were issuing a refund. But while we were doing that, we held up all these other guests. Could we actually have moved? That might be.

Jeremy Julian: 21:46

did we run outta stock between the time they ordered it and the time that, that you had to fulfill the order? there's so many different opportunities, but back to your point, once they know where the data problem is, they can operationally go solve it, rather than being in crisis management of, I'm gonna get in trouble, or I'm not gonna hit my bonus because my wait times are too

Brock Weeks: 22:02

Yeah.'cause before it was just, oh, it's red and it's, they've been here too long. Ah, what do I do? And that was all you were looking at was the fire where it's no, let's, and how do these stages compare cross locations? Because when you're looking at that holistic view, you can really start to go and coach and say, Hey, is this a company-wide process problem? Is this just a store manager problem and training problem? Is this a shift lead problem? It really allows you to get super.

Jeremy Julian: 22:28

Is this a menu mix problem for that store? do they order too many salads and they take too long to produce out of this store versus burgers at another store? Yeah, so many different ways that would take hours and really data analyst tools and people sitting there analyzing these things. What's another use case, Brock, that people are using this for, that you think would blow people's minds to go. How, how have we been doing this so manually and now at our fingertips in an aggregated way? We can just knock it out pretty quickly.

Brock Weeks: 22:55

Yeah. before diving into that, I wanna mention do the same thing on get measuring guest speed of service in the store. Why do we only care about a certain segment of the customer? let's track so that we know if the drive through's home and an in store is not, we want both. the second.

Jeremy Julian: 23:10

It'd be interesting to actually ask that question even on third party, because I see the third party problem being a huge issue.'cause so many brands weren't designed with third party in mind and they've got the regular guests and then they've got the third party guests and they oftentimes will screw up a guest experience because I've got my DoorDash drivers waiting in the same space. sorry. I'll let

Brock Weeks: 23:27

No, and you can track that now, right? Because they typically are picking up at different areas and the AI is good enough. Now it uses a Reid technology that basically says, Hey, I'm not gonna facial analyze Brock, but I'm gonna say he's in a maroon shirt. He's right this big. He's the, he's got this color hair, all these attributes. And then I know what areas of the store he went into. So I can say, Hey. I've got all these DoorDash people just show me the people that just followed this flow. That's why we call it customer flow ai, this flow. What was their experience and how long were they in the different areas?'cause we have, it's, and I empathize with these frontline employees where you're in there and I've got my family and I'm trying to order in store. No one else is in line, but they're just back there humming on third party delivery orders. What I'm

Jeremy Julian: 24:12

And the DoorDash driver's sitting there with his phone going, I need my order. I need my order. And you're like, you're trying to get your kids to go through the line ordering and it's a bad guest experience for both sides, right?

Brock Weeks: 24:21

So it's okay, now let's measure this. Let's identify it, see exactly what the baselines, and now we made a change. How did it impact? And the second piece, back to your, what other technology? we call it site check ai, but historically, when you had to. Train ai. It was really just a computer vision model. It wasn't ai, it was machine learning. And it was, I'm gonna train it on seeing a maroon shirt,'cause that's your uniform. So I'm gonna train this data set and it's gonna get really good at looking at maroon shirts. now AI is just getting to where it looks at an image and says, Hey, I've been trained on all of these large models. What is it you want me to pick out? And so you think about the things that a secret shopper or a district manager does when they're in the store from a cleanliness standpoint. Is table turn happening? Is the floor is being cleaned? Have we put out the floor mats from a safety standpoint? Are the employees wearing their uniform are personal self

Jeremy Julian: 25:10

line checks? Are they making sure that they're doing the temp checks? All of those kind of things.

Brock Weeks: 25:14

the two stopped?

Jeremy Julian: 25:15

many things that continue to happen. Sorry.

Brock Weeks: 25:17

No, you're on the, do you just start to go. Wow, there's a lot there. And though it's gonna continue to even get better right now, think about it, if I visually could stand there from the angle of the camera and look at it like, is it gonna be able to read the serial number off of a dollar bill? No. But is it gonna be able to tell you like, Hey, you got a bunch of tables that are sitting there with trash on there and no guest and they've been sitting there forever. Yeah. Spills, employee uniform, personal cell phone device, safety issues like boxes, covering entrances, things of that nature. But it's gonna happen all day at right at intervals of time because the affordability's not quite there to run it. 24 hours a day yet, but you can say, Hey, every 20 minutes at periodic times, run these checks and just give me a compliance score and then quickly show me where I'm at compliance. So I can look and say, you know what? I'm at 97%, I'm good. No one's perfect. Or, Hey, I'm at 70%. Why? And it will quickly take you right to that point in time where you just hover over the red mark and it's here's the image in the video of what happened. So you're not saying once again, oh, why am I having this compliance? It's, oh, look what's happening. And the reason why we focused on those two key products and there's more coming. It was when they started to look at the customer speed of service. The customer flow started to increase, meaning they weren't getting as many people through. They were decreasing their throughput. There was other problems going on in the restaurant.

Jeremy Julian: 26:45

Yep.

Brock Weeks: 26:45

and so they're correlated together. And so if you can affect, once again, those are two of the core pillar things that affect negative reviews at its core. It now allows you, when you're in the store to not spend time researching, but spend time coaching and mentoring.

Jeremy Julian: 27:01

Yeah. those two use cases are incredible and I love that you chatted through, through that the power of the software is a pretty big piece, I guess Brock,'cause you can go to Costco, you can go to Best Buy and go buy cameras. you talked about. I guess I'd love for you to talk to our listeners about those people that think they have it all figured out.'cause they've, gone and bought them at Best Buy and they're sitting on their couch at night, while their wife's watching The Bachelorette, reading through, watching through all of these videos on their phone. But really at the end of the day, that's a huge waste of time.'cause now software's sitting on top of these things. For those people that kind of went super consumer what. where are the opportunities and what is the delta even in the price of the hardware and the software to get to something that's more professional grade and more commercial grade, to be able to do what it is that they need to do and save them so much time and so much

Brock Weeks: 27:47

Yeah, it's so much closer than it used to be. I do need to point out Savvy's completely hardware agnostic. To deploy our system, you plug in a little device smaller than my cell phone, and that's it. It just compresses and encrypts the data and sends it to the cloud.

Jeremy Julian: 28:02

So you can use any IP camera.

Brock Weeks: 28:03

Any IP camera, we don't care what the cameras, as long as it's commercial grade. Back to that. In order for our analytics to run the camera needs to be continuing to run the delta between a consumer grade camera and a commercial grade camera is probably less than 20% now, like it's

Jeremy Julian: 28:20

Percent.

Brock Weeks: 28:21

hundred percent. It's so close. Like fantastic brands like AWA and Access and Vivo, taking the, they're building phenomenal cameras at really affordable prices, especially for this industry. Where you can go and do that, and there's so many, great resources out there and MSPs and integrators that are offering these at really affordable rates. The key thing becomes is, what's the, what does the technology need to do, which is it needs to capture the image so it needs good light management and good resolution. Do. Do you need eight K at a restaurant? Looking at a lobby? No. Go get just a good camera that has good light management and is gonna last and is reliable once you're there, we say like the magic should happen in the software because it's gonna change. Technology's gonna advance. Just if I had, I have an iPhone here. I wanted to use Apple Intelligence, but I didn't. I had the 14 pro, and so it, it wasn't capable on my hardware. If you approach everything, there's certain things that hardware and software should be 100% married together, but if you do that, you limit the ecosystem that you can play in. The power of what I do with my iPhone is, yes, it's enabled by the processing power, but it is the creativity that has been unlocked from the App Store that I can then download and use from a software standpoint. And so our recommendation is find open systems. You should, if savvy was not performant, you shouldn't be locked to me and be out thousands and thousands of dollars. You should be able to pivot as an organization fairly quickly and so good.

Jeremy Julian: 29:56

So I am wholeheartedly agree. We talk about the freedom of choice all the time at CBS and it's just, it's so critical.'cause again, it puts the onus, onus on us to do it right. And like you, it sounds like you're leading the team in that way. Brock, for those that are sitting here going, you know what? I need this now. What does it even look like? Can you use existing systems and layer this stuff on? Do they need to go run cables to everywhere that they need these things? Can you start small and grow, I guess help walk through what it would look like if people are sitting there going, I need this in my life. I haven't had this. How do I get there?

Brock Weeks: 30:28

Yeah, Where you have 10 stores, a hundred stores really look at it and say, okay, we do a virtual site audit, which is looking at the, what the current camera models are and what the angles of view are. Are they just, do they need to be repositioned? Are they in the right spot or do they need to be added? From there, we literally drop ship our device. You plug it in on the network, it needs to be on the same VLAN as the camera sensors, and if it's integrated to the security sensors and other things, needs to be on that same vlan. N it encrypts, sends the data to our cloud, goes and says, Hey, you need to give me the credentials. I'm in charge now. Send the data. And then right through most of the forward thinking point of sale systems and great ones out there like you all as well. It's just a backend API integration. You hit the button, the data starts flowing, you map the cameras. And honestly from saying, I would love to do this, if everything's in good shape and the hardware works there, you're up and running in a couple weeks. Like the actual time on site is anywhere from 10 minutes to, two hours if they need to adjust angles of cameras. And most of that's just climbing up and down a ladder and looking at the angle adjustment, right?

Jeremy Julian: 31:35

Yeah. two things real quick there. You, we didn't really, dig deep into kind of the other technologies that are in the store and integrating those. You alluded to point of sale, is that your guys's primary, I guess talk through the use case on why would you integrate with point of sale, and what does that do for you?

Brock Weeks: 31:50

Yeah. Point of sale labor systems are a really important one, so you can understand like that ratio of speed of service to labor. unfortunately employee time theft does exist, and so you wanna be able to audit and have the system run and verify those checks, the alarm sensors and systems smart safes. Think about what are the key pieces of technology. That something's happening and interacting with that, I would want to know. Why I got that result, right? Discrepancy in cash deposit, something rang up at the transaction and the way we integrate it there is it's actually we put it all in the cloud and flatten it out so you can search your video data by it. So whether it's a loss prevention use case of saying, show me every time that a cash refund was done that a guest wasn't present, right? That's build an audit workflow, the exception based workflow, it just throws it in a quick digest for you. You click through it in a few seconds and answer the questions. But your product team, and menu team, they might wanna look through and be like, Hey, show me who's buying this item, right? Or, show me when there's a customer complaint comp, or

Jeremy Julian: 32:58

Yeah, gimme every transaction that was over 150 bucks.'cause did we go out and do a table touch when it was a high value customer? whatever.

Brock Weeks: 33:05

All of those things, there's questions around it. Really, it's just allowing you to get to that point in time with the context.'cause when you're watching, you wanna have full context, so with the full context so that you can actually look at it. Diagnose the problem and then go take action. That's how it exists today. Where I see it going, Jeremy, is what gets really exciting. I don't think in a few years people are really gonna look at video. What's gonna happen is you're gonna have these AI agents, right now we're doing this, the structure, all the data together, aggregate it, so checks Abby's doing that. Second, we're applying specific models to specific use cases. Speed of service, guest experience. That's site check. That's almost like a filtering layer and the integrations to the sensors and all that are filtering layers. But when this happened and I saw a discrepancy, then an agent goes and watches the video, analyzes it and says, no, you're good. This was right. Or No, this was potentially bad. Surfaces it into, here's your checklist of action items you need to take. If you wanna click on the link, it's gonna take the same way that AI works right now.

Jeremy Julian: 34:07

You would now, but it, but you know what, you'll have peace of mind that says, you know what, I put my 40 or 50 or 60 use cases in there and it's really only delivering me the two or three. That might be a problem that are coaching opportunities or whatever else I.

Brock Weeks: 34:19

Yeah, those high impact things that you want it to be doing, and it's right, it doesn't replace anyone. What it does is amplify people. To spend less time doing the mundane and more time connecting with their teams and connecting with, Those people that they want to coach, that they wanna mentor, that they, From a loss prevention standpoint was just having lunch yesterday with a franchisee who, once again, why I love the hospitality industry. He was just all about his people. He's got 60, almost 60 stores, right? And he's just all about the team and how he helps, and we were talking about loss prevention and he was like, yeah, I used to look at it as big brother. Like I don't wanna be that. And, but he's I've had the. times over my career where people have stolen significant amounts and we've had to prosecute and why did, what if you caught that person the very first time they did it and just said, Hey, this is a pro. Like you can't do this. This is theft.'cause a lot of times it doesn't start out that way. It's just some innocent discounts, some innocent gifts, some friends, some things catch'em that time instead of having to prosecute and end up with something on their record and really disrupt their lives. It's all of these different use cases, which is really just. Highlight good behaviors, replicate those, coach on them, and find those corrective behaviors so that you can address them.

Jeremy Julian: 35:30

I love that you went there'cause that was my next line of questioning is what do you say to those people that are like, I don't need Big Brother, I don't need all of this stuff. I don't, all of the privacy things. I guess it sounds like you guys have already it ultimately again goes back to how do I make my business better? How do I make my team better? How do I make my customer experience better? Is really where you guys are fo focusing so much of your time and energy in the tools that you guys are building. Is that, is that correct?

Brock Weeks: 35:53

A hundred percent. that's where the impact is at. If I, simple analogy in my life, my big brother definitely hasn't been there to point out and, crack this, the whip at me, he's been there to correct me on, Hey, you might be getting a little off here. I'm seeing this right? let's bring you back. that's truly what it should be used for in our.

Jeremy Julian: 36:10

Love it. so Brock, how do people learn more? How do people get, you talked about, from time to, to even knowing what we're doing to getting up and running it, it can be done in as little as a couple of weeks. So how would they learn more? How would they get in touch? How would they move forward? because again, I. I look at what you guys are doing and I'm just, I was blown away. I was so excited when you shared it with me. I was like running around I was like a little kid, on Christmas morning. Super excited about all of where things have gone because my history really goes back to some of the old days when I last looked at this tech. And so it's amazing how far it's come. for our listeners out there that are excited and wanna learn more, how do they do that?

Brock Weeks: 36:43

Yeah, we're, I'm personally active on LinkedIn. Would love to, chat with you and introduce you to some members of our team who can even, dive in deeper with you. get savvy.com and it's SAVI, is our website. We'll also post a link, in here that has a specific one for this, this show. And then at the different conferences, we're gonna be at QSR, evolution, the Fast Casual Executive Conference, FS Tech, and a bunch of others coming up here this fall. Would love to chance to interact with you all.

Jeremy Julian: 37:10

Awesome. Brock, I genuinely, I was blown away with some of the stuff that you shared with me just in in a brief demo. And so I think, and I say this to listeners all the time, if you're not doing it, guess who is your neighbor that's beaten you to your left and your neighbor that's beaten you to your right is using this technology as a business driver to change the conversation. And so I would really encourage all of you guys to go check out, get Savvy. check out Brock's stuff. they do an amazing job of helping you learn how to get to this place. And so thank you for taking the time, Brock, to share. To our listeners, guys, we know that you guys got lots of choices, so thanks for hanging out. If you haven't already subscribed, please do so on your favorite podcast, tool and or YouTube and make it a great day.

Brock Weeks: 37:50

Awesome. Thanks Jeremy.

Speaker 2: 37:53

Thanks for listening to The Restaurant Technology Guys podcast. Visit restaurant technology guys.com for tips, industry insights, and more to help you run your restaurant better.