AI's Surprising Role in Telematics Revealed

Full transcript

Detecting car crashes fast

So welcome back to the podcast, guys. Today, we are joined by Arnaud Lucas. Why don't we start by you telling us a little bit about yourself, Arnaud? Sure. So my name is Arnaud. As you can hear from my accent, I'm French originally, but at least I've been in US now for twenty eight years. First eight years in Texas doing startups, and then I moved to Massachusetts in the Boston area. I've been doing different thing in my career, but but at heart, I'm a software engineer. And now I'm a I'm a leader in technology. So working in different companies, you know, some are more known than others like TripAdvisor or Wayfair. Some are less known, like, you know, my previous company was a very small stuff like I want to. I was a steo. And currently, I've been at Cambridge Mobile Telematics for more than a year now, which is very interesting company because it's a mission driven company to make the roads and drive us safer. Obviously, So something that they would issue, in my opinion, should care about. So, Arnold, what kind of trends are you observing in the insurance and telematics space

right now? I think so there is a lot of aspects. One of the trend which is not surprising because it impacts every every industry is the use of AI privacy and how to best incorporate AI. So AI is a strong tool. And the question has been how to best incorporate AI into, you know, telematics, obviously, but also, like, the insurance industry in general. And good news for us at CMT is, like, we have been doing AI before AI was cool kind of thing. So company is 17 years old, and effectively what we sell as a company to our customers is not really any kind of data per se. It's not about driving data or anything like that. It's more about, you know, insights on the driving data, be able to come up with insights about how self drivers are, you know, how drivers are involved in a crash are the whole self. And these are kind of insights that obviously we have been using AI and so a lot of machine learning early on. And now we have been creating this thing called TFM, telematic financial model. Think of what it does, but for telematics.

And so I've been creating that so that we can really come up with these new insights in a more in a broader way. So instead of having somebody specifically say, you know what? I think that, you know, for for fun motion, so the fact that users are using their phone, you know, we should check if the screen is on or if the phone is moving in a weird direction or things like that. That's one way, and that's human making that an important. But maybe, I don't know, LLM will say, you know what? If if the care has a support person and something else is happening, then it must be that the user is using their phone, for example. Yeah. They can come up with new insights that we can even think about. So I think that's what's very exciting about from a from a pure thematic standpoint. The and then, obviously, how to make, how to use AI even more on our side too so that we can develop products and and and faster and better effective in general, not just its insights, but everything we do.

AI first approach

So something that has been very top of mind for me. From a insurance business standpoint, like, even end users, like you and me that have to buy, like, car insurance or whatever, are going to start using AI more, you know, to decide which car insurance to go to to figure out how to deal with this car insurance to know? Yeah. Oh, I got into an accident. The first thing I'm gonna do is I pass the AI. What should I do now? So I think there is a lot of things now that AI is part of lot of what people are doing. And then it becomes, well, how do we fit into that too? Right? So that how do we help customers, end users where they are and how that's evolving and where do you plug in to help them with those things too. So I think there are lots of AI has been something that the the thing about AI, right, you know, people say it's another tool. It is. It is a tool. But it's a broad tool that, you know, you and me can use that in our personal lives.

We can use that at work. We can use that now business. We can use it everywhere. Or it's I think that's the impact of AI is the the fact that it's so broad. Mhmm. You know, all that. Now trying to assess each and every piece to see where where where AI can make sense is, you know, is a is is an interesting challenge. That's a good news. Mhmm. Where do you actually use AI in personal life? I'm curious. So I've been using AI for different things. So for example, I'm not a musician. It's funny because everything in my family is a musician, but I'm not. But I've been using AI to create songs in different way, so I've been doing that. So that has been interesting, Johnny. And it does AI has been getting much better at creating, like, songs and music and lyrics and all kind of things. So very interesting. That's that's one way. I've been using AI as well to double check on my son's homework or certain things where he's not sure about, for example. Actually, it's basically things I don't remember very well from my time.

So so there's been a lot of doing, like, very advanced chemistry. I was like, has been a while. So when you're before I do memory a bit. So I have a lot of discovery as well. It can just be because, you know, I want to buy something. Like, for example, I need to I believe that I need a mesh node for my for my Wi Fi in certain area of my house. And I was asking, so what makes more sense? And they said, oh, well, need that kind of your AI mesh node that has at least this capability. Otherwise, you're going to based on the layout of your house, you're going to have a plan or or it's going to reduce your bandwidth. So all kind of things specifically, like, you know, mostly from personal use is more like discovery, meaning that you'll be able to discover new information easily and and quickly. I was I like to do baking. And so in part of

Tech foundation and customer needs

I don't choose AI for coming up with recipes yet. I'm not I'm not sure how it was done. But for example, I was out of bread flour, and my cookie recipe needed bread flour. So I was, yeah, okay. How do I substitute bread flour? And so and so trying to and I have an idea of what I believe I should be doing, but, see, it's like, okay. Give me the exact measurement based on the well, based on what the recipe needs. I think, oh, yeah. Can be used all all over the place. But no. Not Okay. I'm curious what pulled you into Cambridge Telematics. Cambridge Telematics has been very interesting for multiple reasons, and that's why I joined. One is that it's a mission driven company, and it always feel good to be able to work for a company that wants to save lives and has a proven record of saving lives, which is even better. But it's not that they want to. They actually are saving lives. So I think that that's a great company to work for. Obviously, the company is profitable and doing well, which is always a good environment,

then, you know, otherwise so that that works. The other aspect too that's interesting about the company is the people. You know, for me, I really wanted to find a company where I believe I can grow. And the way I can grow is by working with, obviously, smart people that know what they are doing. Sure. But although that are not jerk and actually they want to they are here to help you and collaborate with you. And and I believe that we are exactly that. We have a very inclusive company where people are nice, humble, and yet very, very, very smart. So so that's for me, that's the perfect combination to learn from, to grow, to get feedback, to to do all the right things so I can actually do my best. And so that for me, that was the most important criteria to best when I was looking for a new opportunity. I feel like of that or that definitely exceeded my expectations. And then finally, obviously, the technology. Like, I'm working on my areas called mobile sensing intelligence. There's a combination of mobile with sensor data,

and you'll be able to use AI effectively on the phone to determine things. For example, you know, let's say you get into a crash. Right, you don't want to wait for the data to make their way to the back end, and from the back end they say, how are you in the crash? Too bad. And then and then you're dead, right? So obviously you want to the front to detect the crash directly to say, oh, need help. Okay. Great. Get help right away and not wait for for things to make their way through some neighbor's cloud somewhere. And so that's part of what what they're on, specifically. So to enable that, so that way that's one way to to save lives, if we do. Is there something that you've changed your mind about in the past year? It can be in regards to either personal life or business or something specific in your industry. I'll go back to AI because I do believe that AI is progressing very rapidly. And I do believe that no. Too bad. I've been using I've been using LLMs now for for three years.

Yeah. Three years. You know, almost as soon as it came up. And I've always found the tool very useful. I do believe for the last couple of months now, we're only seeing a step change function. Like, feel like to be propagandized, like last year, you know, it almost felt like technology had reached some kind of plateau to us. Well, I felt like, okay. It's useful, but it's not getting that much better. I think, yeah, I want the edges. I thought she's just not telling me I have two a's in blueberries anymore. But still, like, it's not it's not fantastic. Like, it's not like, oh, now I can just tell you how to do anything for me, and I don't have to check. I'm not sure we are there yet, but, well, the latest versions of of some of the models, especially at work where I've been using it, they you know, where I've been switching in mass from cursor to countdown, has been very relevant. Well, you could say, okay. This is definitely getting there. And so I think that's part of

the change in my mind. That's why I was like, okay. Well, yeah, I can be a good helper, you know, on the side. And when when you're not sure, you kind of ask or you ask for help, and you you kind of talk it through with AI in different way. Well, now I'm like, no. That's you know, there there is this push with I believe in now, which is no. It sure must be like AI first thing. Like, you want to do something. Check with AI. Or, you know, try to come up with a discover what you need to know to know, plan further what you what you want to do, and then, you know, execute on that effectively. And every step of the way, I truly believe that now AI is at the point that you can help. Mhmm. As is help me. So that does kind shift in what what changed for me. I don't know. If you could give one practical advice to someone in the same industry, so telemetrics, the software space. Right? What would that advice be? I'm not sure anymore because I used to say that

one thing I got very lucky about is not because now you could argue, okay. I'm a leader. I'm a manager or another. But I started to to be a very integral contributor, a very hands on software engineer for a very long time. And I believe actually, I truly believe that was the right thing That you want to hone your craft. You want to have strong fundamental that you use lot of expanses to to build upon. And and that's, you know, no matter what you do after, at least you have that base. That's not going to go away. Yeah. Technology changes all the time, and you have to adjust to that as part of being in this industry. But at least you have that foundation you can kind of count on as as your best. That's something that, you know, anchors you. Joel, it's unclear to me how AI will change that, right, because before I say, well, yeah, if it doesn't work to be a manager, I can always go back to coding and that's cool. Well, coding now is like, well, yeah, I can do it for you and I can't. So

so it's going to be interesting a bit. It's true, really, in a strong foundation because ultimately, you know, yes, AI could replace the coding, but there are a couple of things. For example, when you're trying to architecture a solution that I don't believe, for example, the AI is there. There's a thing about thinking about how it better fits into the business and what you're trying to accomplish from a what kind of impact you want for customers and so on. I don't think the AI is there either. So I do believe that there is opportunity by having that strong technical foundation to then bridge that technical foundation with to business with your customer empathy and other thing. And I think that, you know, somebody was telling me recently that that's why I know Amazon is talking about having builders more than just software engineers, meaning that, you know, this mix between product managers and software engineers and designers altogether, but the point that your customer has a need, you know, you figure out the need, you implement a solution. Great. You know, now the customer out there needs met in some shape or form.

Do believe that requires a strong foundation across the board in technology and the understanding of what makes sense and how you can help your customers. But, yeah, that's that's we'll see how my advice leaves the times of of AI. I don't know yet. Well, yeah, definitely having a strong foundation helps. That's that will be the truth even with AI, I would say, that, you know, having strong foundations, strong skills will be always good. I I I do agree still. I do believe in that. I also believe that, you know, I think AI showed us more than ever that it's important to adjust, stay flexible, stay curious because that foundation we may still have that foundation. It may still be important, but you may exercise it in a very different way. So you don't want to get stuck on something. I remember one of my neighbor, and I was like, so what are you doing? He said, oh, I was a software engineer, but I'm now I'm retired. And it's like, so why I wanna try finding a job? Well, I only can do Cobalt, but there's no job for Cobalt anymore. I was like,

okay. I mean, learn the new language. Right? But I think that that's that's the point, especially in our industry. We have budgets for sure. Mhmm. And just being adaptable Mhmm. To what's happening in the world. Okay. Thank you, Arnaud, for joining Thank you. Today's podcast. I will add links, and thank you guys for watching. Thank you.