Mike Chung: 0:08

Welcome to AutoCare OnAir, a candid podcast for a curious industry. I'm Mike Chung, senior Director of Market Intelligence at the AutoCare Association, and this is Indicators, where we identify and explore data that will help you monitor and forecast industry performance. This includes global economic data, industry indicators and new data that will help you monitor and forecast industry performance. This includes global economic data, industry indicators and new data sources. Hello and welcome to another episode of Indicators, part of the AutoCare OnAir podcast series. I'm Mike Chung, I lead market intelligence at AutoCare Association and I'm very happy to have Alex Roblitz, the founder of Demo, with me. So, alex, welcome to the show. Mike, thanks so much for having me on. Really a pleasure to have you here today. So, alex, maybe tell us a little bit about yourself, your role and what Demo does Sure thing.

Alex Rawitz: 1:02

I've spent over a decade in tech startups across the telco, finance and now the automotive world. I've been working on Demo for almost five years with my co-founders, and our mission at Demo and a lot of what I've always tried to be a part of and the other tech companies I've been with is how do we get more out of data. That's a really really, really big question to try to go and answer, and our take on that at Demo and the problem we've been focused on specifically is how do we help people everyday drivers, you know, fleet managers how do we help them make their data more accessible? How do we help them be more in control of their data and, basically, how do we help leverage all kinds of new tools that are emerging in order to build better products.

Mike Chung: 1:54

Thanks for that. Introduction Really gives us a good springboard to jump from and I think you're on the right episode to talk data. So, thinking about automotive data, I think you have what 280,000 or so vehicles that you're getting data from?

Alex Rawitz: 2:08

Yeah, almost 200,000. And so when we started Demo, we really set out to build two paths. One is to create an open ecosystem and a platform that makes it easy for developers and enterprises and companies to have easier access to vehicle data. And two, we built our own mobile app, which is available to pretty much anyone in the US, europe, canada and a few other countries coming online soon, and it's a telematics-based mobile app that allows you to track and store privately the data about your vehicle or vehicles. And so over the last three or so years, we've added almost 200,000 vehicles across almost 100,000 individuals who are now privately storing their data.

Alex Rawitz: 2:57

So what kind of data? It's a great question. A lot of the data has started with the basics of telemetry data, so we're talking about where you're driving, how fast you're driving, the engine data, the DTCs that are coming from the vehicle, all the stuff that is readily readable from the vehicle canvas. We also let users add other kinds of data to this little data vault that they have that can include things like their title and registration, condition photos and condition reports, service history. We're sort of building a digital twin for the car that everyone can be in control of you said ETCs earlier Diagnostic Trouble Code, so DTCs.

Mike Chung: 3:43

Oh, dtcs Okay. Layer Diagnostic trouble code, so DTCs Okay. So I think of telematics, data, some of the things you highlighted, certainly location speed. I think of other things like heartbreaking. Maybe give me a couple of more examples of the type of data that are being collected.

Alex Rawitz: 4:00

So I'll do that, and I'll do that from the perspective of really who our users are today.

Alex Rawitz: 4:05

Our users, our customers, are people who are pretty data driven individuals.

Alex Rawitz: 4:11

They're the type who probably has a smart home and is monitoring maybe their thermostat or other devices in their home. They might be the type to have a wearable who's tracking their health and their fitness data, and it's somebody who is finding comfort or opportunities in actually looking and exploring their own data. So some of the things that we display in our app today include also, like your fuel consumption, and from that you can actually start to create a bunch of derivatives of that data, whether it's predicting how much fuel whether it's gas or electricity your car consumes, what your expected spend rate will be over the year. We can help track efficiency of either the battery or the engine itself in miles per gallon. We can see tire pressure. We can detect issues that could be affecting the car's performance and efficiency. So our goal is to actually expose as much of that data to the user and expose trends and insights so that, as an individual, you can feel a little bit more in touch and maybe like you're taking a little bit better care of your car.

Mike Chung: 5:17

I appreciate hearing that Thinking of the individual consumer first, and then we'll go to fleet managers, because you alluded to that earlier. When I'm driving my car and I think about the data that are on the dashboard or when I hit the menu button, certainly more is there, but I have a tendency to only really study it while I'm in the car. It sounds like this allows somebody who's, as you mentioned, data-driven and David-oriented, wants to improve operations and efficiency.

Alex Rawitz: 5:47

It allows him or and over the life cycle of a car it adds up pretty quickly If you take out depreciation, the average, the costs going into an average vehicle every year about $8,000 per vehicle.

Alex Rawitz: 6:17

And so small things, small inefficiencies or unaddressed issues can start to add up in your fuel costs, in your long-term maintenance costs, and that was something that, for me, I was concerned about in my vehicle and was wondering what could I not know about?

Alex Rawitz: 6:36

That's happening very slowly. That might exert some long-term effect on either how much I'm paying to just use my car every day, or even in the residual value of the vehicle itself when one day I want to sell that car. So for us it really comes back to a lot in terms of how do we empower users and how do we empower individuals to kind of have better control and better insight into that, and understanding the data is one part of it, but of course then you have to be able to do something with that data. So a really big part of what we've built in an area that we're investing in more and more is how do you actually make that data shareable? Could I share that data with a repair shop or an insurance company or an AI even, and use those tools to find further discounts, further savings, further ways to just be a better car owner.

Mike Chung: 7:26

So maybe an example could be this I'm subscribed, I collect the data and perhaps I'm looking at the data and is it something where perhaps there's an AI solution that'll say Mike, your fuel efficiency is X, Mike, your fuel efficiency is X, and if you did not accelerate while you're going downhill or you use less braking, based on your vehicle records, the maintenance that you've been doing you can save X number of dollars on fewer brake pad changes by changing your driving in such a way. Are those types of things possible? That is possible.

Alex Rawitz: 8:04

It probably was hard to do a little while ago, but you know, now, in an era of great analytical tools and, of course, great AI products, we can do that and we're doing some of that already today in our app both passive and prompt-driven AI experiences. So we will have, within our app, you would see the AI telling you about certain things. Maybe it's how to interpret a check engine light that has popped up. It could be even recommendations, like the one you mentioned on how to get more fuel efficiency. Or you can prompt the AI that's built into our product, which has access to the vehicle data, to say you know what I feel like? My fuel efficiency is going down. Is that something you can confirm, or can you speculate as to why that might be?

Alex Rawitz: 8:55

And it's possible that the agent could look at your throttle position and say well, look, you've really been, you know, pressing the throttle a little bit too heavy and that's going to destroy your efficiency. You need to just ease up as you press the throttle and these things sound, I think you know. Maybe to some people they sound a little bit like they could be small, little impacts. Who's really going to be? Who's going to care that much? But the reality is our data and our user base and the people that we speak with. There are millions of people that are looking to make these changes and these adjustments and in ways that they can actually just improve their ownership experience.

Mike Chung: 9:36

I appreciate hearing that and touching on something you highlighted earlier in terms of the consumer opting in. I know that that's very important from a data protection anonymity. Don't sell my data, so it sounds like that's up to the user in terms of how much data he or she is willing to have go into an AI application or go to their insurance company or a repair shop. Is that fair to say?

Alex Rawitz: 10:02

Yeah, we believe that. You know we think of cars and this isn't by any means a demo invention in terminology but cars are computers on wheels now, yeah, the same way, we have a computer in our pocket and when you connect to the internet from your phone and go and browse websites, you have a lot of choice as a consumer as to how much data you want to be available. You can go on the web and go on Facebook, and you can happily use the internet and know, hey, all of big tech is scraping my data. We're all used to that narrative, we're familiar, we know it's happening. If you want to get a VPN, if you want to browse privately, there are browsers and VPNs that you can do that with.

Alex Rawitz: 10:48

There are steps you can take to protect your privacy online. You can deny cookies. There's all these choices you have when you're browsing the internet, but that hasn't made its way into this other really important computer in our lives our cars. You basically have the option to say I want a connected car and I have no idea where the data goes, or I want a dumb car because I'm worried about where my data goes if I have a connected car.

Alex Rawitz: 11:14

And so what we've done at Demo is really try to bring that same level of choice and optionality that already exists in the wide worldwide web today to this new frontier of connected vehicles. So, as a user of demo, if you just want to use your own private instance of the app, look at your own data, take better care of your car entirely your choice. If you're one of those people who says, cool, I'm going to share my data with my mechanic and my dealer and this person and that person Great, you can do that too. You could even say, hey, I'd love to sell my data to the highest bidder. If I can get paid for that, like, why not, who cares? Those are the choices that we think should exist.

Mike Chung: 11:53

Fascinating and thinking about the data. I know that a lot of telematics solution providers the data is coming from their phone, but it sounds like you're connecting to the car itself. Is that correct, especially for like the diagnostic codes and so forth?

Alex Rawitz: 12:11

And phone-based telematics is super common. I mean, it's how many of the apps that you know and love probably work today, whether it's all the mapping applications and navigation apps. The only problem with phone-based telematics as it relates to the auto world is you might use my data when I'm sitting in an Uber, and there have been these reported instances where the phone-based telematics I'm not the person driving or I'm on a roller coaster. I think there are some crazy story like that where those non-driving instances have even affected people's rates and things, and so we spend a lot of time talking to data people at insurance companies and automakers and many, many other types of companies, and there's sort of this tacit universal acknowledgement. I should say that the phone-based telematics is really great because of how broadly it applies, but it does have its very well-known drawbacks.

Mike Chung: 13:06

So in terms of people that want to engage with DEMA. Do they have to have a vehicle of a certain that's not over a certain age, and things like that?

Alex Rawitz: 13:18

Yeah, the age today is roughly we support vehicles roughly back to 2008. We do that with an OBD2 device. We have a couple of great partners that have been building those devices. Of course, today we're at a really kind of we're at an inflection point in the industry. 95% of new cars are built with telematics built right in, although many automakers are still in the early days of figuring out how much data they can make available and how much they can send and what they're willing to pay for and things like that. So we're at an inflection point and we're starting to see, of course, more and more that consumers are demanding better experiences with their connected vehicles. So we expect that the data component, even though 10 years ago was the first date, is the new oil conversations, or 15, 20 years ago. We're really getting to that part of the industry in the automotive world today.

Mike Chung: 14:16

So you mentioned insurance companies, you mentioned fleet operators. Just thinking about your clientele, what can you tell me about the type of data that they're looking at, their use cases?

Alex Rawitz: 14:28

Yeah, it's a really great question. We sort of think of the companies that are interested in using vehicle data. In a couple of buckets there's kind of your everyday usage items and these are going to be the things that most people are thinking about on a day-to-day basis maintenance related. So you know, those are simple. Factors can be what kind of car do you have, what's the fin, what's your odometer, how much are you driving? Do you have any codes? Pretty simple amounts of data.

Alex Rawitz: 15:01

In-car experiences are starting to leverage vehicle data more and more. I don't know if you saw, but Ford launched a Zoom integration. You can do Zoom calls in your car now and those applications need some access into understanding what's going on inside the vehicle so that you can join a Zoom call when parked versus when not parked. So there's kind of a lot of data about the status of the vehicle Are you parked or you know those sorts of things about the status of the vehicle Are you parked or you know those sorts of things.

Alex Rawitz: 15:27

Another category of data is or really use case is like these big ticket items insurance selling your vehicle, trading it in, whatever it might be. Typically those are going to want a bigger package of data. An insurance company might want to see a month of driving data to give you a telematics-based quote, and a vehicle marketplace might want to see a month of driving data to give you a telematics-based quote. A vehicle marketplace might want to see a whole history of data. They maybe don't care where you're driving, but they might want to know what kind of issues and data popped up along the last few years of ownership.

Alex Rawitz: 15:59

And the last category that we talk about is new digital experiences, and this is where things get really fun and interesting. We're even working with video game developers who want to try to build applications games based on vehicle data. A lot of times it's going to be geo-based, so they're looking for location data, but they're thinking about really creative ways of using speed to influence. You know how good of a player or role you have in this video game. There's really a lot of really cool emerging use cases that we're excited to see come out over the next couple of years.

Mike Chung: 16:34

That's fascinating, thinking about the fleet managers, for instance. I think we touched on some of the use cases there. So it could be mileage, it could be acceleration, braking, length of trip, the maintenance related variables, whether it's has the oil filter been changed, things like that. Are there other things that fleet managers are engaging with?

Alex Rawitz: 17:06

engaging with. You've certainly covered the big ones. There we see fleet, a whole wide range of things. I mean. I've talked to fleets that are interested. Maybe they're a bus fleet and it's how many people are on the bus they're interested in. There's a lot of interest in predictive maintenance these days, so a lot of people are hoping to kind of just pump in. Maybe every can signal that you can get and hope that an AI can tease out some expectations. There's also, of course, you know, a huge component of how data is used is often in the resale of vehicles or the consignment in some cases in large fleets. Can we prove the health of the vehicle when we take it to auction so that you know maybe we can avoid an arbitration at the end of the auction or something like that. There's, you know, honestly, tons. I mean the fleet use cases are a little bit less varied than I think we'll see on the consumer side, but everything is trending towards automation and AI Gotcha.

Mike Chung: 18:04

And you mentioned that in your organization you have basically a data analytics team. Am I remembering that correctly? So what kind of interesting findings has your team come up with that you're able to share so we've?

Alex Rawitz: 18:16

taken a number of different views of data and there have been a few topics that we've looked at. One of our very first areas of interest was around EV really around range anxiety, I should say and we looked at the data of course, you know, anonymized and all that from users in our app who are driving internal combustion engine vehicles gas vehicles and we imagined that they had a level one charger at home and that they had a pretty, you know, basic EV. We said how many, how often would these people really need to charge away from home with a basic EV? And what we found is that the average would be like once every three months. We sort of I don't know in our mind we really debunked the whole range anxiety issue that you know. Most people probably would not spend a lot of time looking for chargers if you have even a level one charger at home. We also found that there are tons of instances and I don't remember the exact data points off the top of my head there are tons of instances of people with pretty low battery capacity in their EV driving past functional chargers that are maybe just the block over from where they're driving and they're not seeing or they're unaware. Maybe they don't have an app that helps them locate that charger and they're missing those opportunities.

Alex Rawitz: 19:44

So we've done a lot of research into the EV space and, from what we can tell, range anxiety is not as scary as it should be and there's a lot of opportunities for people to charge. Something else that we've looked at on the more of the like maintenance side we're we're always curious like where are people going to be spending money and what are the impacts of uh check engine lights? You know issues that people have catalytic converter efficiency and I mentioned earlier in this conversation that a big factor in what we're doing is how do we help save people money? If you are losing efficiency in the catalytic converter, you're going to see your miles per gallon suffer and it's going to be something that could impact the dollars out of your pocket in the long run. So there's a lot of things that maybe don't affect your ability to drive the car today that could of your pocket in the long run. So there's a lot of things that maybe don't affect your ability to drive the car today that could affect your pocket over the long run.

Mike Chung: 20:46

Thanks for sharing those insights. I think the first one's really fascinating, particularly from the what is my normal driving pattern and how often do I stray from it? How often might I go do a long weekend trip or use my car for, like a summer vacation? So I think that that's really interesting to hear as you and your team are looking at the data that are coming in to do these types of analyses and kind of test those null hypotheses I think that's what they used to call them in Stats 101, right, do you find issues with data or kind of like oh, this data is not as clean as I might like it to be, or there's so many makes and models, and how do I streamline it? Can you talk to us about, you know, any kind of hurdles that you've had to clear when dealing with data.

Alex Rawitz: 21:35

There's a couple more than a couple, I should say Some that come to mind. So we work a lot with data that comes from OEM APIs, automaker APIs, and one of the things that we have seen as an issue in some of those APIs is that the data will come in, often in forgetting the exact term, but they'll often come in fractionalized data points and there can be discrepancies between when you receive the data point and when the timestamp really adheres to. You have to sometimes knit disparate data pieces together. They'll come in out of order. Making sense of some of that data is actually often quite complicated and at times we've had to integrate really multiple data streams, sort of like hey, give us the real time data stream, then give us the theoretically cleaner data stream. Be, maybe the issue is the technology under the hood, but sometimes it's also just hey, the car drove out of cell reception and now it's sending the backdated data and it's coming in at order. So there's a lot of that that we have to deal with.

Alex Rawitz: 22:52

When you're plugging in devices into vehicles, as many of our customers do, you do have to have a significant amount of decoding of individual data points on each vehicle. There are a lot of standardized data points across all cars but there's a lot that's proprietary to each automaker and each make and model. So we spend a lot of time rinsing out whether it's anomalies in the data that really shouldn't exist in GPS location or speed or engine position or anything like that. We put a lot of work into trying to rinse out those individual issues. Those are the two biggest ones that come to mind.

Alex Rawitz: 23:34

I think you know probably one more I'll add, just for the sake of it is validating this data is really really hard a lot of times. The best way to do that, of course, is to have your own hands on the vehicle and have your own computers hooked up to it. But when you're trying to do this with a customer who's maybe on the other side of the world, it becomes quite challenging to understand. Hey, are we seeing the same thing? Are we getting the exact same results? What's really going on in the car? Because you need to match up what you're getting in your bits and bytes with you know, the atoms that are actually in the real world.

Mike Chung: 24:11

That's a fascinating point, because these are unique driving events to a unique individual. So can you reproduce that? And to the first point that you made, it even came up when we were talking about recording this podcast. Right, if something kind of blips out for a little bit, it's being recorded locally, but it'll quote unquote catch up. So I wonder if that might be an improvement on OE or API developers' minds to say, oh, we're out of cellular, we're going to just store this locally and then, when we get to a certain point, put it back in.

Alex Rawitz: 24:45

Yeah, and for some use cases, if you're shopping for insurance or you even just have a usage-based insurance program or you're getting your car remotely diagnosed via API or something, does it matter if things are out of order for half a second or, you know, the data is maybe 10 seconds delayed? Probably not. But as we enter a world of autonomous vehicles, where these vehicles need to react and interact, you have to solve that problem of really true real-time data. Of course, any AV needs to be able to handle its environment without connectivity. That's something that every AV company is building into it. But the more real-time and the more redundant those data feeds are, the more opportunity you have and basically, the smoother those experiences should be.

Mike Chung: 25:43

And I think it'll be interesting too, because you talked about that lag where, on the human side of the analysis, you and your team are accounting for that, and I would imagine that, as we go forward, there will be an integration of oh, we've seen this movie before. We have a process to handle the data and clean the data and, as it goes into more of an AI oriented analysis, that will have been learned, I would guess, too.

Alex Rawitz: 26:54

Yeah, there's always this balance between you know, constantly refining how that data is cleaned and analyzed and retaining. You know, one of the things that that always also inspired us at DEMO was thinking about a future where you have two autonomous vehicles and, let's say, they crash and nobody's in them. Who's at fault? How do you determine that? And there are some, of course, that are building their own verticalized stack, but many AVs are. There's a car, there's an automaker and there's a sensor and there's software built here, and there's actually dozens of companies, hundreds of companies, that are part of that vehicle's stack. And so you want clean data at the operational level.

Alex Rawitz: 27:36

But if something goes wrong, you do want to be able to understand what part went wrong and how is that recorded and certified and how do you have some level of trust in which part of the stack is at fault, in some auditability there. So we spend a lot of time thinking about how do you balance? You never really want to throw out bad data points, because bad data points can actually mean something. In some instances they tell you something about what's going on. Instances, they tell you something about what's going on. We're going to live in a very complex world where understanding good data and bad data, but also knowing how those things were produced and how we can verify, either or is really really important.

Mike Chung: 28:18

You bring up some really interesting points, because I think the standard for human error is different than the standard for the machine error. And a simple example could be this when I think about ADAS systems, like lane departure, I'm driving down the road and it's a stormy day and a tree branch falls in the road and I have to swerve to avoid it. If the lane assist is on, will it allow me to do that? And then to your point about AVs. I mean, that's just a very simple example of here is something that can happen in a dynamic world that you and I acknowledge. But you know, have we as programmers thought of every scenario? Is the AI capable of predicting all of these types of things, right? So I think it's hard for me to imagine and as a simple human being, I suppose you know an optimal solution for everything under the sun, right?

Alex Rawitz: 29:21

It's a really interesting. That problem in particular is super interesting to think about Today. Whoever's in the driver's seat, the buck stops with them. Right, you can have the lane keeping on, you can have full self-driving on. I actually happen to use an aftermarket ADAS system called Comma AI, which is really really sweet, but at the end of the day, whoever's sitting in the driver's seat, the buck stops there.

Alex Rawitz: 29:46

But we're already talking to automotive companies, insurance companies that are trying to think about when does that change? At what point, you know, do you have to factor in? Okay, you know, the car made a hard braking or a hard stop, or accelerated or swerved or something, but it was the lane keeping system's fault and not the humans, and it misinterpreted something I don't think that anyone, at least not to my knowledge, has come up with. You know what is the right point at which responsibility flips or becomes shared, or how do we build models and factor that into the future? It's a really really thorny question that, you know, one of the you know, in the very near future we're all going to be tackling.

Mike Chung: 30:37

Right and I think and not to say that I'm an expert on this but I have definitely appreciated kind of the human wisdom and element of, let's just say, that type of incident happens tree falls down, people get in an accident, police officer comes and says yeah, you know, it's hard for either of you to avoid this. I'm not going to assign blame right here, so I think there's having that opportunity for discretion and judgment will be certainly interesting or important. Thinking about things we were talking about earlier, like when I asked you about what have you and your team learned, I can just only imagine the possibilities of analysis, because you get all this data from the vehicles, but you could overlay that with other data like weather data right, that's not necessarily being collected by the cars. So that just opens up the whole, opens the floodgates for all kinds of interesting research, research, exploration, I would think. Is that fair to say?

Alex Rawitz: 31:43

Yeah, we have gotten quite a few requests from people that are, you know, I wouldn't say they're uninterested in the car itself, but they're looking at how that data can inform other types of analysis, whether it's how can we better understand consumer patterns. You know a lot of people like to. We haven't talked to any hedge funds and, by the way, like you know, at Demo we don't share data with anyone. The analysis we do is just internal, but we do have people that come and ask and are interested in that. So there's lots of interesting opportunities there.

Alex Rawitz: 32:21

We are in the early stages of talking with a smart city project. The real estate developer. There's an automotive company, a telecom involved, and what they're interested in doing is basically taking our ability to privately collect and anonymize this data and how, all the vehicle data and patterns that you can draw out of it, and they want to use that to better understand how to plan these smart cities that this real estate developer is building. Where should we be citing electric vehicle chargers? Where is their congestion? Where is their? Where are their opportunities for more parking?

Alex Rawitz: 33:00

I mean, these sound again like simple things, but the average, and there are some tools that exist for this today, certainly at the public, you know, in the public sector, but nobody's really had the ability to say, hey, we want to deploy something within our community, within our development, with our whatever, in order to better understand what's going on, and that's something that that we're actually able to help them with. So I've, I hope that we can do this project in the next couple of months, Cause I think, you know, we have a few baseline things that this, these folks want to collect, but we're really, really excited to see, like, what are the opportunities? What are we going to discover that we had no idea was even going to be possible.

Mike Chung: 33:39

Thanks for sharing that, and earlier you talked about the number of vehicles, the number of households. Is that US only and, if so, can you tell me about? Are you looking to expand your kind of data provider base, if you will, to other parts of the world?

Alex Rawitz: 33:56

So when we launched Demo the app about three years ago, we launched it in the US, canada and Europe and since then we realized that as we want to expand more and more internationally, we need core partners in new markets that help to bring Demo to those markets. So there are two markets that we're really in the early stages of expanding to now. One is Japan, where we have a sister company now that is going to be bringing our technology and our products to folks in Japan, and two is a partner in Latin America. It's actually a very large car dealership group is a partner in Latin America. It's actually a very large car dealership group and they got really excited about what we've built because they can offer two compelling product experiences to their customers. One, they will white label the Demo mobile app and provide it to their customers, which is a really huge benefit in Latin America, because cars are not being made as natively connected it's much, much earlier down in LATAM, so they can offer their customers an app that can help them manage their car. And two, they will have the users assuming the users agree share their data back with the car dealership and so now they can have somebody on the other side who says calls up the user hey, we just saw a check engine light come on P0430.

Alex Rawitz: 35:13

Actually, it's an issue with the catalytic converter. Don't worry, it's not urgent. But hey, we have an appointment for you on Thursday. Would you like to come in at noon? And that's a service that this dealership group is really, really excited about rolling out. They can offer a digital subscription super important to any dealership to get people back into the service base. It's where they make a lot of their money, and so they're just really, really excited about how they can use data to deliver this personalized service, digital service to their customers.

Mike Chung: 35:46

Right, yeah, and it makes the relationship that much stickier in terms of customer loyalty and, like you said, being able to keep them on as a long-term revenue customer, for not just the vehicle but for service itself. Yeah, Yep, Exactly right. So, Alex, so many exciting examples of things that are happening. As you look into the future a year, three years, five years, 10 years what do you see in terms of the types of data, the types of applications that can be that you and your group are going to be exploring?

Alex Rawitz: 36:17

Obviously, you know everyone's talking about it, but for us it's it's all about AI. And how do we and when you talk to companies that are really deep on AI from startups to, you know, the Amazons of the world they're all after one goal how do you make AI proactive? You know most of our socks have been blown off. Collective socks have been blown off in the last few years by things like ChatGPT and Claude and all of the amazing LLMs that exist, but their fundamental flaw is you got to write, you got to type something you know. Sure you can speak to it, you can talk to it, but you're giving it a prompt. And can these AI agents draw enough context from the data that they receive to become unprompted or basically just be prompted by changes in the data you know? Something that I would love is to, in the way we think about it is to make this world possible. You need to have your own private data vault and you can mediate access to that data vault, but as a user, you want to have all your data stored in one place.

Alex Rawitz: 37:35

Things that are annoying or you have to think about to deal with in your car are going to go on autopilot. When do I need to worry about my next oil change? When is my registration due? Again, I forget. Do I have the best insurance I could have, or should I be shopping around? All of these things that most of us don't have a lot of time to consider but would like to answer and have on our agenda or have in our calendar or something like that, should become super automated, and that's something that's going to happen for individuals. But to the point of the car dealership and that relationship and that loyalty, it's probably the biggest single area for companies to start distinguishing themselves from a loyalty and a service perspective, because now you can provide a customized experience and, of course, we're going to need different AI agents to talk to one another, which is something we're also experimenting with. There's so much that's going to change, but it's all about how do you use data to become more proactive.

Mike Chung: 38:29

Fascinating and that can certainly make it more frictionless. Kind of perhaps declutter my mind of, like you said once, my next oil change and kind of allow me to focus on other things, if you will. This has been so fascinating to talk to you, alex, and you know. Just tell me a little bit more about yourself. You mentioned you're a kind of a technology entrepreneur. You've been in a couple of different industries.

Alex Rawitz: 39:18

You've been in a couple of different industries. I I went to some entrepreneurship classes and I was like this is kind of like the homework would be like come up with 30 business ideas to like talk about as like that that's not all that useful. And instead I was like selling ads in the newspaper or like helping to run a moving company locally in college, and so those were the things. That was really where I cut my teeth and I just loved the small environments.

Alex Rawitz: 39:44

By the time I was graduating I couldn't imagine working at a big company. I knew it had to be someplace small where you know you're wearing a bunch of different hats and you're exploring a lot of new things and really you're living and dying by whether you're wearing a bunch of different hats and you're exploring a lot of new things and really you're living and dying by whether you're picking up the phone or getting the delivery done or whatever it is. So I just love being at a Deem was about four and a half years old and almost 25 people, but it still is exciting to be in an early stage environment.

Mike Chung: 40:15

Thanks for sharing that, and I'm thinking about our young listeners, who might be in college or relatively new graduates, and it sounds to me like your experience has proven that when you're in a smaller organization, having your hands in different things, you learn a lot, and perhaps the distance from start to finish for a new idea is much shorter, involves less people and can be allowed for faster innovation.

Alex Rawitz: 40:40

You know I was always a very good student throughout my entire career in, in, in in school and stuff. But honestly, when I look back on all that, I, I, I kind of like I get scared as I think about having my own family like just how much of the like again. Like coming back to this word of prompt, like school is all about like provide a prompt, give an answer, and the real world just isn't that way at all. And so I was really lucky to, in college, kind of break out of that where I realized, hey, the book learning is great, there's certainly a lot of value there, but like there's a totally different skill set instead of learning that you do when you're in the job versus in the classroom. And that was something that. That's something I think about very often and I wish anyone who's kind of going through that right now you know, embraces from an early point that you know it's a different, it's so so, so different.

Mike Chung: 41:38

Really wise words indeed. So thank you for sharing that. And as we wrap up here, thinking about summertime, thinking about vacations, do you have any favorite vacation spots? Or like kind of a favorite dessert you like to have flavor? You know? Just something fun. I guess I know that's a lot of random things, but no, no, no, I'll, I'll.

Alex Rawitz: 42:01

I'll pick up on the the favorite vacation spot. One of the places that I absolutely love that I think about often is Copenhagen. I just think it's such a cool city. I've been there a couple of times for work. I liked it so much I brought my wife there. We had an amazing time. It's just a really uniquely amazing and beautiful city. I think it kind of gets missed when people talk about great places to go in Europe, and it's someplace that I couldn't recommend more.

Mike Chung: 42:31

What about Copenhagen makes it so appealing? Is it the food, the culture, the arts?

Alex Rawitz: 42:36

I think it's a little bit of all those things. I happen to love the Nordic and Scandinavian food I mean, just like a beautiful piece of toast with like the fresh caught shrimp that they have. I mean, if you love seafood, it's an amazing place for fresh seafood there and that's one thing that I really, really love. You know, I think that like there's just a, it has the right balance of activity and kind of laid backness. Now I come from New York City, so basically any other place on earth feels more laid back, but it just like is a really great place. Like if you just want to have like an afternoon beer and do the like European, like sidewalk cafe thing, it's amazing, great to know.

Mike Chung: 43:16

Thanks for sharing that with us, and is there anything else you'd like to touch on regarding data before we wrap it up?

Alex Rawitz: 43:23

Yeah, you know I was so excited to come on to this podcast because, as I mentioned towards the beginning, you know our mission is to help give people more control of their data and we believe in privacy. We believe in people of their data and we believe in privacy. We believe in people owning their data and you know that's something that we feel really strongly about and we're trying to introduce to individuals and to enterprises, and so, you know, I think that's going to become more and more important. Of course, we're seeing in the EU Data Act and then the Right to Repair Act here in the US. But you know, just the last plug. You know, if you are interested in Demo, you can head to our website, democo, and learn a little bit more.

Mike Chung: 43:58

Fantastic. Alex, thanks so much for joining us and to all of our listeners. Thank you for tuning in and make sure you smash that like and subscribe button and share this episode with your friends. So, Alex, thanks for having us and thanks to all of our listeners for joining us. Until the next time, have a great day. Thanks for tuning in to another episode of Auto Care On Air. Make sure to subscribe to our podcast so that you never miss an episode. Don't forget to leave us a rating and review. It helps others discover our show. Auto Care On Air is proud to be a production of the Auto Care Association, dedicated to advancing the auto care industry and supporting professionals like you. To learn more about the association and its initiatives, visit autocareorg.