Lance Glinn:

Welcome into another episode of the Inside the ICE House podcast. Today's guest is Daniel Ramot. He is the co-founder and CEO of Via. Daniel, thanks so much for joining us inside the ICE house. Happy to have you here.

Daniel Ramot:

Thanks. Real pleasure to be here.

Lance Glinn:

So take us back to the very beginning before Via existed. What did you see in transportation systems that felt, I guess, fundamentally broken or inefficient, and what were the sort of key gaps in the market that you wanted to address with your team and create this company?

Daniel Ramot:

I think the key thing that we saw, my co-founder, Oren Shoval and I when we started Via, was a gap in the fundamental concept of how public transit works. So if you think about public trains and historically transit, it really depends on buses and trains. These are large vehicles that run in fixed routes. In the case of rail, you would have to even build it out. And for their success, they really depend on very large numbers of users all traveling in the same direction, basically at the same time. That works great. We're here in New York City. Amazing.

Lance Glinn:

I take NJ Transit every single day.

Daniel Ramot:

I take the subway pretty much everywhere I go. The subway's amazing. Some of the buses in Manhattan, other parts in New York are amazing. Very busy. If you're going down 5th Avenue, this is a great solution.

Most parts of the United States in particular and other parts of the world look very different from a major city. There's not as much density we've chosen to build in a very spread out way. So if you try to run a bus through those neighborhoods, you end up with what we have in a lot of parts of the United States today, which are very inefficient buses. Nobody's riding them. And then, as a result, you can't afford to ride them very frequently, so you run them even less frequently, once an hour perhaps, and then even fewer people want to run them.

So we had this thought that, well, could we use smaller vehicles, shuttles, think of them vans, minivans, and run them in a way that their routes are constantly changing. So totally dynamic routes, have people summon them on their phones. And in that sense, be able to create a transit system that's much more dynamic, data driven, technology enabled, and add to the bus and the train systems that were there.

The question was, well, could you really make that work as a mass transit system or would it look like a cab? And after we did a lot of work, we came to the conclusion it could really work as a mass transit system.

Lance Glinn:

Yeah. And clearly has. I mean, Via has had great success, obviously, enough to join the ranks not long ago of our great listed companies here at the New York Stock Exchange. And when you dug deeper into those challenges and those gaps in the marketplace that you were trying to solve, whether it was the fixed routes, or underutilized capacity, or lack of flexibility, how did you begin to conceptualize a model that could better match supply and demand in real time?

Daniel Ramot:

So the first thing we needed was data and there wasn't a lot of data out there in the beginning to say, "Okay, how do we design these algorithms? How do we make decisions?" When someone opens up their app and says, "I want to go from here to there." How do we figure out which vehicle to match them with, which route that vehicle should follow, and so forth?

And so we realized that a few years before we started, New York City had installed these credit card readers in taxi cabs, in yellow cabs. And in order to make sure that people weren't cheating on the fare, they tracked the GPS location that the fare was done right so of the pickup and the drop-off so that they can compare to the fare. And so they collected, at the time, there were about half a million yellow taxi trips in New York every day. So over years, they'd collected many, many millions, tens of millions of trips where they had the data of the pickup and the drop-off.

And so we submitted a Freedom of Information Act to the city and asked for that data. And I believe they had 30 days to respond to our request. And on day 30, we got a bunch of DVDs.

Lance Glinn:

They waited all the way up to the wire.

Daniel Ramot:

Yeah, it took a while.

Lance Glinn:

We kept you on your toes for as long as they could.

Daniel Ramot:

Kept us on our toes. And we got a bunch of DVDs in the mail, which was pretty cool, with all this data. And so we had millions and millions of pickups and drop-offs across New York City. And so we started to analyze, well, okay, if instead of taking the taxis, we created this dynamically routed mass transit system, would it work? What would the routes need to be? How many people could we serve? And we discovered that we could serve that demand with a lot fewer vehicles and taxis that were available and that you would get a very large amount of sharing, essentially, of people in the same vehicle at the same time going in the same direction. We launched a service to try to experiment with that. We can talk more about it. But that was really the beginning and the way that we developed the original set of algorithms.

Lance Glinn:

And I think some, and I think rightfully so, some would look at transportation just blindly as like, "Hey, this car, train, bus, plane, whatever is going to get me from point A to point B and that's sort of what I need it for." But how do you think about the broader value of public transit, whether it be to people or to cities beyond just getting from point A to point B?

Daniel Ramot:

So I think, especially here in the US, we have gotten used to the private cars, the primary way that people get around. Today, about 90% of people in the US commute by private car across the country. It's sort of fundamentally our mode of transportation.

The challenges that it creates are, number one, a lot of congestion because you have a lot of people sitting in their own cars. It's a huge amount of cost that comes out of that. And you might say, "Well, you know what? It's still the most efficient way to get around given the way our cities are laid out." But I think what's starting to happen, especially in the last few years, is that the cost of owning a car is becoming very, very difficult for a lot of families to support and sustain. Today, the average car buyer is 50 years old. Younger people just cannot afford to buy... This is for a new car.

Lance Glinn:

Sure.

Daniel Ramot:

For a used car, the average repair is close to $1,000. So if your car breaks down, you need to really ask yourself, "Will I be able to pay rent if I have to spend $1,000 on this repair?" And if that car is the only way that you can get around, then you might lose your job if you don't have a car. So public transit fills a critical need for many, many people. And I think probably a lot of listeners of this podcast may not really feel it emotionally, but it is the case. It is a critical way to encourage employment, education, to support education, and really enable people to get around. Not to mention that we have a lot of people who also have certain disabilities and special needs and they also rely on public transportation, but public transit becomes really... So those are the people who absolutely need to use the system and we have to provide them a solution.

I think public transit becomes so much more when we get what are called choice riders into the system. So people who have a choice, who could otherwise ride a car. I think that's what we have here in New York. If you look at the subway, there are lots of people who could probably afford to drive, take a taxi and Uber, but they choose to take the subway or the bus. And that's when I think we create great cities.

So if you look at some of the greatest cities in the world, one thing that's common to them is they all have great public transportation. And so I think public transit is... There's a social element to it, there's an economic element to it. We've seen some really nice studies. There's a recent MIT study shows that, in Chicago, for every dollar invested in public transit, you get $11 back in economic activity that it stimulates. So there's so many ways in which public transit is crucial.

Lance Glinn:

And you mentioned in your answer, too, that there are people that have different needs, right? The transit systems don't necessarily take that into account. They're built already. Some are, especially I'll take NJ Transit for account, something I take every day. Some of those trains are old. They're old, they're outdated, and they fit a certain customer, but not all customers are what they were built to fit. How do you think about being flexible and making sure that what Via provides fits a different type of consumer? Because you might have someone who's a senior citizen, you might have a young child, you might have someone with a type of disability, you might have someone that is just normally walking down the street. So how do you make sure that you adjust to what the different needs and the different expectations are for every customer?

Daniel Ramot:

I think that's a critical question and one that makes some of these newer modes of transit that we are supporting, or existing ones with maybe better software and better technology, so important because you do have... People are very diverse in their needs, their capabilities, their financial ability. So being able to provide this range of solutions to them is very important.

So having vehicles that are accessible to people in wheelchairs, having an app that is accessible to people with low vision or no vision, supporting people who don't have access to technology. You may not have a credit card. We take having a credit card for granted perhaps, but there are a lot of people who don't have a credit card. How do they pay for that trip? How do they reserve it? They should be able to call. Maybe that's their only way to access.

Some folks are very good at booking a trip when they need to go, so on demand or jumping onto a train or a bus. Other people really need to have a certain level of certainty that that trip is going to pick them up, so they may need to reserve in advance. So creating a system that has all the flexibility to address all these use cases I think is critical and is also part of the social contract that I think we have with residents of towns and rural areas to support them.

Lance Glinn:

And I had mentioned NJ Transit and I had mentioned there's a lot of history, obviously, behind it, not just NJ Transit, but all of the rail systems, especially around here in New York City. There is a lot of history with them, but I think there's also a lot of history of sort of fragmentation between transit or train service, buses, cars. How do you dream or think of sort of an optimized, one size fits all type of system for all modes of transportation, rather than having them be separated and sort of individualized?

Daniel Ramot:

So now you're really talking about what I dream about.

Lance Glinn:

Okay. Let's get into it.

Daniel Ramot:

So let's get into it.

Lance Glinn:

Let's get into it.

Daniel Ramot:

I think Via started with, as I described earlier, with microtransit. So that's what these dynamically router shuttles are called. They're called microtransit, that's the name that they've been given.

That is one piece of the transit system. There is another piece which is what's called para-transit in the industry, transportation of people with disabilities. That's a pretty big part of the system. In the US, we spend about $7 billion a year on para-transit services, take people who are in a wheelchair or need to get to a dialysis appointment and so forth to wherever they need to go. And then, of course, you've got the buses, and beyond that, the trains.

Today, these systems typically operate in silos, so separate operations teams, separate software, separate payment systems. What we really want to do is bring it all onto one platform. So you as a rider can open up your app and say... You don't really care. Is it a bus? Is it a microtransit system? You just want to go from point A to point B.

Lance Glinn:

Yeah, in the best way possible.

Daniel Ramot:

And you don't need to want to worry about, "Well, how do I pay for this? And then when I get on the bus, how do I pay for that? Do I have change? Do I have a Metro card?" All these silly things that we would have to deal with in the past. You just want to do it all in one seamless way. And from the transit agency side or the city side, you want to ideally manage that entire system in one place. You want to use the same... If your vehicle can be used from microtransit and para-transit, and you want to probably use that in a most efficient way, maybe some of your vehicles can run dynamically some of the day and then start to run fixed routes during other parts of the day where there's demand for them.

And that ability to put everything onto one platform is something that we've been working on at Via for quite a few years now and are starting to increasingly do, to bring together all these disparate modes, to put all the things... When you're a transit agency, you're planning your system, then you're scheduling it, then you're operating it. These have also historically been separate systems that you'd have to transfer data in between very, very inefficiently. So to put all that on one platform adds incredible value to the city.

And today, that's what Via offers, is a single platform on which you can run every mode of transit all the way from planning, to execution, to operations, all the data and analytics that you need to feed back into your planning to change whatever you need to change. And that is adding enormous value to cities around the world.

Lance Glinn:

And just based off your answer, it obviously shows that that is something, that optimization, that bringing everything together is something that you're really passionate about. So when you were starting Via, when you were building it up, when you go public, were you surprised that there wasn't something like that already in existence? Because I would think this optimization, bringing everything together, is something that's so obvious.

Daniel Ramot:

It seems obvious.

Lance Glinn:

Obviously, as times change, when it was first built, when NJ Transit was first built, when Metro North was first built, it made sense in a certain perspective to have them all individualized and all separated. But in the world we live in today, I would think for a city like New York City, for example, having everything under one app, or one payment system, or one schedule system would just be common sense, no?

Daniel Ramot:

Yes. You would think that.

Lance Glinn:

I would think if I were you, building up, I'm like, "Why hasn't this existed?"

Daniel Ramot:

So I continue to be surprised by some of the things we're discovering as we get deeper and deeper into these transit systems. Now, historically, if you look at these systems, they have not had a ton of technology embedded in them anyways. And then these departments have grown over time to operate in separate ways. And it's just sort of the reality of many of these transit systems, the way that they currently operate. And change with government is risky. As a government official, whether it's an elected official, someone who works for a government agency, you're often not rewarded for taking a risk, for driving change. And so you're really, in some sense, incentivized to just kind of stay with what you have. You know it works a certain way.

Lance Glinn:

It's one of those, if it ain't broke, don't fix it type of things.

Daniel Ramot:

Right. And if you do try to fix it and you get it wrong, there's a real risk to that.

Lance Glinn:

It backfires.

Daniel Ramot:

So I think that changes is not very fast in this industry, but we're starting to see it accelerate. One thing that I would say about this industry is that when you have one city adopt a solution, the cities nearby notice, and the residents notice, and that starts to create both pressure to drive the change. "Why aren't we operating that way?" And also reduces the risk for the buyer on the other side, for the government agency. And so we're starting to see in certain locations, certain states, certain regions, a real acceleration of adoption of these new solutions, which for us is pretty cool and I think, most importantly, for the residents who rely on these systems.

Lance Glinn:

And so on a more personal note away from Via for a second, what was your own founder journey like? Because I know you were in the IDF, we talked about growing up in Israel in the IDF, eventually deciding to build a company focusing on transportation and transportation services. What was just that process like to figure out, "Hey, this is what I wanted to do and I was going to start pursuing it."

Daniel Ramot:

Yeah. I mean, in our case, Oren and I, we knew we just wanted to start a company together.

Lance Glinn:

Sure.

Daniel Ramot:

We wanted to be entrepreneurs. We didn't know what it was about. We spent about a year... Oren was doing his PhD. He was at the time at Harvard, I was working here in New York, and we would meet somewhere in New Haven. We'd take the train up and down and we'd try to brainstorm a bunch of different ideas. Had a few pretty terrible ideas along the way.

Lance Glinn:

Do you ever go back and wonder, if you could think back to those original conversations, because you say you want to start a company, you don't know what that company you want to start is, right?

Daniel Ramot:

We had no idea.

Lance Glinn:

It could be in any service, right? Do you ever think back now that you've created something that's super successful and be like, "Hmm, I wonder if we had decided to go this route what life would be like. Or that route, what life would be like?"

Daniel Ramot:

Sometimes, probably more so in the difficult moments where you're thinking, "Oh, what the heck am I doing right now? I wish I'd done this other thing."

Lance Glinn:

Sure, sure, yeah. Sure.

Daniel Ramot:

But yeah, we really didn't have a clear direction, but over time we came up with three criteria for ourselves. And we said, one, what are we good at? And we felt that we were good at solving difficult math and computer science problems that applied to the real world. So we wanted to do something that had sort of a physical manifestation. In our case, vehicles, drivers, riders, people using the systems in the real world. But behind it, we wanted there to be a real computer science kind of challenge, which I think the system, and we can talk more about it, the system that we've built solves some really interesting problems, computationally and mathematically.

The second thing we wanted to do is we wanted to identify an industry where, if we were successful, it could become a very large business, which I continue to believe this is the opportunity to grow even much larger than where we are today. And the third thing is we wanted to find some area where if we were successful, again, we would have a positive impact on society, on people. And those were our three criteria. And we weren't sure we'd come up with an idea that sort of met all those three, but that's what we were looking for.

At some point, Oren had this idea that... In Israel, we have a class of public transit called sherut taxis. These are vans, they're very low tech, so they're just vans that run along fixed routes, but they're smaller vehicles. You typically flag them down anywhere along the route and you jump in, you historically would pay with cash, you'd sort of pass your cash up to the front, you get the change back. And Oren had this idea that we could put these on an app, basically, so you could reserve a seat, so you wouldn't have any uncertainty when you got up to the road. And then we started to talk about, well, okay, if you could reserve a seat and you knew where you were and where you wanted to go, maybe we could then redesign the route in real time based on those requests. And that's sort of how the idea began to grow out of that initial concept.

Lance Glinn:

With that experience of those sherut taxis, did you ever run into an issue, and you said it earlier in the conversation too, that people want certainty, right? They want the ability to know that this vehicle is going to come at this time, there's going to be room for me, and it's going to get me from point A to point B at the time I need it to get me to wherever I'm going. With the sherut taxis when you were in Israel, did you ever run into the issue of, let's say you flag one down and there just not being space, there not being room or if something happened -

Daniel Ramot:

That was exactly the root of the idea. So I think from Oren's stories, he would say he was trying to catch one of those, and he was standing along the road and they kept coming by, but there was never a seat available. So he said, "Well, what if I could book a seat on my phone? Then I'd know I'd have a seat. That would give me some certainty." And then, well, like I said, if we have this information, maybe we don't have to follow these fixed routes. You could just tell us where you are and where you want to go and we'll create a route for you. And if we created a route for you, could we create it for... But not as a taxi, or at the time Uber was just starting, there was Gett taxi. There wasn't even a Lyft back then. What could we do then to create this dynamically routed, shared system that is much more like a mass train system than a taxi?

Lance Glinn:

And what were those early challenges like, and how much conviction did you have, and how often was your conviction challenged during those early times? I feel like all these founders that I speak to, there were always those instances where they run into a problem and they're like, "Maybe this is just the end of the line -"

Daniel Ramot:

"It was a terrible idea."

Lance Glinn:

"- For this idea and I might have to start a new one." But ultimately they find a way to then get through said challenge and create something that is globally known, domestically known, and really successful. How difficult was it for you in those early days, and Oren as well, to deal with some of those challenges? How much conviction did you have to have to continue to press on even if obstacles came?

Daniel Ramot:

Yeah, we had some moments, for sure. I think the key moment of question was when we said, "Okay, we're going to build the system." We got the data from New York City. We had these algorithms. We were running these simulations that looked very compelling as to how efficient of a system you could create using these dynamically routed shuttles. And then we said, "Well, we've solved public transit."

Lance Glinn:

It's done. We're done. They're amazing.

Daniel Ramot:

And we're going to just call up the MTA and they're going to say, "Where have you been all our lives? Why didn't you come sooner?" And so we started to try to get in touch with these transit agencies and cities and say, "Listen, we think we have something really interesting here. Would you like to work with us? Would you like to pilot it? " That was sort of our original idea.

And we literally could not get a meeting. When people understood that what we wanted to talk to them in transit was dynamically routed transit, they either canceled the meeting, hung up the phone. People thought this was the dumbest idea they'd ever heard. And they would say, "You guys don't understand. What you're doing is never going to work. And if it does work, then the routes that you'll end up running you'll see are basically the bus routes, because we know where people want to travel so we put buses where people want to travel. You're not going to discover anything that we don't already know."

And we were looking at the data that we had and we were thinking, "We don't think that's right. We're seeing people who are traveling along routes that don't look anything like the bus system and those routes could be very efficient as far as there are lots of people trying to travel along them." And so at that moment we had no customers. It seemed like maybe this idea was not going to work. And that's when we decided we were going to launch our own, essentially, ride-sharing service here in New York City and we were just going to try to show that this works ourselves. And that was another moment where we struggled, because then we said, "Well, what vehicles are we going to use? We need large vehicles."

Lance Glinn:

How do you get the infrastructure to back it up?

Daniel Ramot:

That's right. And so Uber, of course, was already running at the time as Uber Black and they were using this model with... It became Transportation Network Company and so forth. So there was a model that we could do that was asset light, so that was one great thing, but which tried to fit multiple people in this sedan that doesn't really work for mass transit system. So then we were walking around the city and we spotted Suburbans, and we thought, "Well, could we turn these Suburbans into effectively buses, into dynamically routed shuttles?" A Suburban is one of the most expensive vehicles out there as far as this transportation. So we took an $80,000 vehicle and we made it into a little bus, and that was kind of another interesting step along the way.

Lance Glinn:

And so you've brought up, obviously, these dynamic routes quite often in the course of our conversation so far. And I find it interesting, and I'm curious sort of like the, I guess algorithm behind it, because let's say you're picking me up in at the New York Stock Exchange, right? I go with Via, I say, "Pick me up at the New York Stock Exchange." And then there's someone else who's trying to get the same ride, but they're 25 minutes away, completely off the route that I want to go to, or very far from where I ultimately want to end up. How do you and how does your technology sort of figure out, "Pick up this person, don't pick up this person, go this way, don't go that way." Right? How does that all work behind the scenes?

Daniel Ramot:

So that's sort of one of the foundational algorithms that we have today is that decision making. You request a trip from the New York Stock Exchange to wherever you're going, somebody else requesting a trip on the Upper East Side going downtown. You probably probably can't match you in the same vehicle. It doesn't really make any sense. So there's a question of how many vehicles do we need at any one point on the road in order to meet all the requests? How far are you willing to go out of your way to pick up somebody else? Probably depends on how much or how little you're paying.

And then we started to discover that what we really needed to do in order to solve this problem were a few things. One, we couldn't really pick you up... We can't probably come to the New York Stock Exchange exactly. We actually probably want to ask you to walk a little bit, maybe half a block, a block, to the corner, so that our vehicles can stay along the main routes. That turned out to be pretty key. So we created this concept of what we call a virtual bus stop. Virtual bus stops can be pretty much on any corner in a sense.

Lance Glinn:

And I think most people would be willing to walk the block to get to a corner if needed.

Daniel Ramot:

That was our intuition. Our intuition, too, was when we think about our own experience trying to catch a cab, or even if maybe with an Uber, you would maybe walk to the corner to just make it more efficient for yourself. You might say... So if you're going to join a shared ride, walking to the corner is not that big a deal. So that was one thing.

The second thing that we found out was that we needed to understand the demand patterns very well so that we could predict where we... So we don't want to be surprised by your request here at the New York Stock Exchange. We want to know that there's a very good chance at a certain time someone is going to be requesting that trip. Of course, you need to get to a certain scale in order for that to work.

But once you get to that scale and you can predict where people are going to be requesting these trips, you can send the vehicles in ways, along routes, that are pre-positioning them to capture that demand. If you don't do that, it turns out it's very hard to be very efficient. So that's the second key thing.

And the third key thing was people are very funny in the sense that we could take you up, say the West Side highway, wherever you're going, Even if it's not the fastest route, you might be very happy with that because you feel like you're going in the right direction. If we start to take a different route that may be faster, but let's say it goes in the beginning in the wrong direction, people really don't like going in the wrong direction. They immediately think that we're doing it because we're trying to pick somebody else up, we don't have their best interests at heart. There's a lot of psychology in how you perceive your trip that turned out to be really important. People, for example, don't mind making detours to pick other people up at the beginning of their trip. As you get closer to your destination, you get increasingly irritable about these things.

Lance Glinn:

Yeah. Anticipation of arriving, I think, heightens when you get closer to your destination. And I also think too, and something you did mention, I know when I'm driving, I will take a route that is, let's say, five minutes slower or gets me to my destination five minutes later, as long as I have time, if it means that I am, say, going on a highway that doesn't have a lot of lights, rather than, say, going on a route that might be five minutes faster, but that has a lot of lights, has to make a lot of turns, has various bumpy roads or whatever. I'll take that five minutes extra if it means it's a smoother drive as well, which I'm sure I'm not the only one who probably thinks that way either.

You want the most convenient drive, like you said. You want to always be going in the right direction rather than having to go backwards to then go forwards. I find that all that information is just so... The psychology that you sort of had to think about in creating this brand and this company.

Daniel Ramot:

That's right. And I think you're exactly right. The intuition is that there is the fastest route, but then there's the route that you would perceive as a good route, recognizing that you're willing to pick up other people along the way. You understand the system that you're getting into, but you have a threshold beyond which you're like, "This was really dumb. I'm not going to do this again." And so we started to measure, well, for every trip you take, what is the probability that you would be willing to come back, essentially? That you will request another trip and how do we impact that probability by changing the route, by changing the price, by doing all that? And so for several years, we really studied, in some sense, human behavior when it comes to how people perceive their travel.

Lance Glinn:

I think there is a fastest route, obviously, for a human, but I think there's also a best route for the human, and it seems like you really were looking for that best route. Obviously, you don't want the fastest route to be 20 minutes more than the best route, but trying to find that in between from the fastest route and the best route seems like something that you were trying to do.

I want to stay sort of on the topic of technology, but I want to focus on technology more with the assets. And there's been a lot of conversation right now around autonomous vehicles. It's something that a lot of people are thinking about. When you look at the current landscape of the technology of autonomous vehicles as a whole, how do you personally think about its level of maturity and what interest Via has in potentially adding it to its portfolio?

Daniel Ramot:

So for us, autonomous vehicles are a really interesting and key topic. My experience, at least taking, for example, Waymo, is that the technology has made enormous progress over the last couple years, to the point where it's both remarkable, and then once you get in the car, but once you get used to, in some sense, unremarkable, as they like to say too. And so it feels like finally we're at a stage, after many years of promises that haven't necessarily been met, where it's really coming, this technology's really coming.

Now, I think the challenge is that the cost is probably still very high to run these vehicles. So realistically, unless you are a company that is developing one of these systems and willing to invest in it, actually utilizing it in a public transit context is probably still, for the very near term, at an experimental stage, pilot sort of stage. I want to pilot here, I want to learn about it.

But our expectation is that, over time, the cost curve will come down very significantly, to the point where maybe, in certain circumstances, maybe in all circumstances, it's much cheaper to use an autonomous vehicle. Now, they have other advantages. At least based on the reported data, they also seem to be safer. So would you be willing to pay a little bit more perhaps to have a safer system, perhaps? So those are things to consider, but it definitely feels like they're coming. How quickly that cost curve will come down so that it makes sense to use them at scale across public transit, that's very hard for me to say, but maybe within a few years.

So the way that I think about it is we partner with cities and transit agencies, we provide the network layer, the layer that just does everything we were just discussing, orchestrates the matching passengers with vehicles, dealing with dispatch, planning the system and so forth. The driver can be a human or can be a robot. From our perspective, in some sense, from a technology perspective, it shouldn't really matter. We're just orchestrating that network. And over time, as the cost curve comes down, you could imagine being able to provide more transit for a given budget by using autonomous vehicles.

That's a pretty exciting opportunity, especially because cost is so important in public transit. You have very limited budgets, you're trying to do a lot with those limited budgets, those services really matter to people, they don't have a lot of money to pay. And so this could really accelerate, in my view, public transit. And we're looking to partner with all of these autonomous vehicle developers, and that's what we've been doing, to deploy those vehicles into our system so that we're ready to go as soon as the cost curve starts to come down.

Lance Glinn:

Absolutely. And I think all of that points to both company and society transformation, company and society evolution. And I think with that transformation and evolution ultimately comes, in just one word, growth, right? Continued growth, continuing to sort of not necessarily change, but become more mature, try new things, experiment new things, obviously see what works. And as you look when it comes to growth, Via has expanded across the country, including in my home state of New Jersey, I looked recently too, with a partnership with NJ Transit. Whether it's new geographies, new service models, or potentially different types of fleets, how do you determine the right areas of growth and sort of the right geographies to then further pursue?

Daniel Ramot:

So we're operating today in about 40 countries. Our biggest market is the US, but we have a very strong presence in Europe as well. And our core focus is to grow today within those geographies. We have 21 states where we consider to be in what we call a flywheel motion. So that's motion where we've started to see this acceleration that I described earlier. Enough cities and transit agencies have adopted the solution that as they look, as other cities are looking across the municipal border or whatever it is, they're seeing these opportunities, and that's accelerating our growth without having to invest as much in sales and marketing.

These geographies where we're already starting to see this very positive kind of reinforcement cycle. And within those, we're pushing very hard, obviously, to grow across all those modes that I described earlier, and our key focus is on trying to take over entire transit systems. So where we've seen really, really nice success is when we're able to deploy our system across the entire public transit system. So buses, para-transit, and then adding the microtransit layer that we bring. That allows us Just to analyze the system and say, "Okay, you've been running the same bus routes for decades, but maybe the population has shifted, where the place of employment are has changed. People are working from home post COVID. The whole dynamic may have evolved in a way that these bus routes may not necessarily be the most efficient."

"Number two, you've run them because you haven't had a choice. There hasn't been any alternative to the buses, but maybe you're running them in these suburban neighborhoods, going around in circles, these very circuitous routes once an hour and very few people are riding them. That's not the best use of funding. Let's remove those bus routes. Let's replace them with the microtransit solution. Let's run buses only where you really need them and then try to run them as frequently as humanly possible given your budget. And then we'll bring people to them with the microtransit system and create what are turning out to be much more efficient transit systems for a given budget."

That's very hard. That change is very hard to drive when we're only in a piece of the transit system. We can add a microtransit system, we can help modernize the para-transit, but what we really want to do is take over that entire system, analyze it in that way. [Inaudible 00:30:58]

Lance Glinn:

And that speaks to the optimization that we talked about earlier.

Daniel Ramot:

Exactly.

Lance Glinn:

Bringing everything under one silo or one network, versus the fragmented ones that are currently in place.

Daniel Ramot:

Definitely. So bring everything on one silo, utilize the resources in the most efficient way. And that delivers a lot of return for the taxpayers, frankly, who are paying for these services and then allows us to grow even further. So this is, for us, a really interesting growth area and one that I think is delivering a lot of value to the communities.

Lance Glinn:

And so we talked about autonomous vehicles. We're talking about growth right now. And I ask this to a lot of the CEOs that I have sit in that chair, because so many represent so many different industries, so I get a lot of different answers. How do you think about, in the field of transportation and transportation services, how do you think about sort of staying ahead? Because it is a field that is evolving. It is a field that isn't staying stagnant. How do you think about staying ahead so that what customers of Via will want next, you will already know and be prepared for when ultimately they do want it?

Daniel Ramot:

Yeah. So maybe I'll answer in a couple parts. First, to understand our customers, the current state of their technology and what's available to them as far as what people are... Very few entrepreneurs, unfortunately, companies with sophisticated technology, are going into the business of selling to municipal governments.

When we first started, before we did our IPO in the New York Stock Exchange, we started a long process of educating investors, public investors about who we are. And the first set of meetings were always about, "I'm sorry, you sell to whom? And they buy... Who are these entities?" These cities that a lot of times folks haven't even heard of can be very large customers for us. Sioux Falls, South Dakota, Evansville, India. These are amazing cities that you wouldn't necessarily think of as having these very large transit budgets, but they do.

And those customers, unfortunately, have not had great access to technology. And so just providing them the most modern tech stack that allows them to integrate all their operations, handle things like payroll for drivers in efficient ways. Things like that are a huge step forward for these customers.

Now, we don't want to just deliver them software that was best in class last year. We want to deliver them the most cutting edge products, and that's where AI I think comes in. So how do we bring AI to these customers? Because a typical dispatcher sitting in a dispatch center running a train system is probably not vibe coding her own dispatch solution. So how do we bring AI to these folks and help them leverage the best possible technology?

And I think from ours, where we sit, Via has the potential to benefit hugely from AI because we can be the conduit for these customers, for these municipal governments to access AI, where otherwise it just seems like something they're hearing about mentioned in the news and very few other companies, if any, are really offering them those kinds of solutions. So I think that's a really interesting area for us to drive progress. And the other thing that AI does, which we can talk more about, is it allows us, given we have these really strong relationship, trusting relationships with governments to ask, "What else can we offer them that goes beyond transit?"

Lance Glinn:

Yeah. And I'm sorry to interrupt. I'm glad you brought that up because I know Via recently launched its new AI labs to help expand its capabilities and see, as you just said, what else these governments and the customers that you have want outside of what Via is currently offering to them. So what was that strategic thinking behind this AI labs to, again, expand the capabilities and potentially, and maybe I'm wrong about this, but I would think potentially branch out from what Via is currently known for?

Daniel Ramot:

Yeah. So it's a great area that we're very focused on. We've launched this initiative, as you said, Lance, AI at Via labs, and we do think of it as a lab where we're developing new solutions for our customers who otherwise don't get a ton of access to these technologies.

What we've seen over the years is that we're working with the municipality, they have a lot of challenges that could be addressed very effectively by modern technology, but building that for them, historically, would've been a huge engineering effort. And also cities are very funny and they wanted a product to look exactly the way that they want it to look.

Lance Glinn:

They want to customize.

Daniel Ramot:

They want to customize. So let's say if you go out to buy a TV, you're not sitting at home speccing the TV that you would like and then issuing a document saying, "Okay, who can build me this TV that I just designed in my head?" You go out and see what's in the market, you buy the best TV or best car.

Government procurement works very differently. Typically, people get together and then write up what they think the system should do. Now, of course, they understand to some extent what's out there in the market, but sometimes they don't even care. They say -

Lance Glinn:

They're dead set in their ways.

Daniel Ramot:

"I want this system." And therefore, selling to these customers requires a lot of flexibility in the system. So either you're building custom solutions, which we at Via decided that we're not going to do, or you're building one tech stack, but then it has to be heavily configurable, has to have a lot of capabilities. This is your typical sort of vertical tech. And to do that in another vertical outside of transit, when we've looked at it historically, has been a huge engineering investment.

Today, with AI, we can build things much faster and we can make them agentic, which means that that tool kind of learns all these customizations that you want. We don't have to tell it in advance what it needs to do. So we can build in these new verticals in one city and then take it to other cities very rapidly. So we're deploying the forward deployed engineer model to work with several cities now, pretty actively, on a few really interesting problems that they're trying to solve, where they're just very, very manually working through these huge burdens on the cities, that we can build tools for them very quickly that help them save enormous amounts of time. And our thinking is that we can do it fast with AI and we can take that solution and deploy it across other cities very quickly. It's early days, but it seems very promising. It would be a meaningful expansion, obviously, of our TAM.

Lance Glinn:

How hard is it to... You have customers that know Via for what it was when it was started. And when they think of Via, they think of transportation services, they think of this microtransit that we've been talking about, this para-transit, et cetera, optimizing everything into one thing. All the things we've talked about, that's what they know Via for. How challenging is it to then go to them and say, "This is what Via has been. This is what Via can still be for you, but hey, we also have these new AI labs that we're adjusting or that we're building up. What else do you need?" How difficult is that conversation to introduce them to something new? Are they receptive to it?

Daniel Ramot:

I think you're hitting on a really key question. How hard is it going to be to convince them that we can do other things for them? And it's certainly a barrier. I would say it's a much lower barrier, though, than for a new provider to come and try to sell into these cities. It isn't so easy to create these relationships. And I think where we have these existing relationships that help us overcome this issue is, typically, we're very close with the most senior decision makers, the mayor, the city manager, the city council. And so we can go to those folks and say, "Hey, you know us. Hopefully you trust us. We've been delivering for you and we've been delivering solutions. We haven't just been selling you, 'Here's the software, just use it.' We've helped you with everything you've needed to do very successfully. What else can we help you with?"

And we found those folks, at those levels, very open to working with us and engaging. So our typical mode of operations, we go into the city and we meet with the mayor and then all of the senior decision makers in the city, and we sit together, the fire chief, the police chief, the county clerk, the city clerk, the city administrator. And we're sitting there... The person responsible for building permits, the person responsible for zoning permits. And we're sitting there really trying to understand sanitation, all of the problems that they have and what we can quickly spin up for them that will help them solve a real problem. And can we do it in a way that we can then take it to other cities to help them? That level of access that we have, I think, is pretty unique.

Lance Glinn:

Yeah. Those relationships obviously matter.

Daniel Ramot:

They really do.

Lance Glinn:

And like you said, having the ability to go to a city and say, "Hey, look at what we've done for you already. You trust us. What we've been doing is working..." Most importantly, because they're not going to keep using you if it's not working. "So it's working. Here's what else we can do for you." I'm sure those already preexisting relationships and that trust certainly will go a long way as you continue to obviously build up this AI lab.

So Daniel, as we get towards, now, the end of our conversation, when you just look forward, when you think about the next five to 10 years, what you want Via to look like, how you want the company to transform, what does success look like in your mind? Is it a number? Is it reaching a certain milestone? Is it something intangible? What is success for Via moving forward?

Daniel Ramot:

So to take a step back for a second, I truly believe that local government has to work well for people. Yes, some folks may be worried about how well our military procurement is working or some strategic geopolitical issue, but their main interaction with government is at the local level. And when local government doesn't work, that's when they start to lose trust in government in general and they start to feel the system isn't working, "The system isn't working for me." And I think that's what leads a lot of people also to the extremes when it comes to politics.

So our fundamental mission is to make local government work better for citizens, for the residents. We've started with transit. Our goal is to have Via system run every public transportation system in the world, I'd say even some countries we don't work in. So in the countries that we do operate in. And then, beyond that, to become that trusted provider for local government, just to make local government work better for its residents and for citizens. I think if we can do that, that drives up trust in government and sort of changes the whole dynamic.

Now, we can't fix everything. It's our small piece, but that's what we're really focused on. We've seen in cities like West Sacramento, where we have a really great microtransit system that's been running, that the city runs a survey around the public's trust in local government, how they rate local government on an annual basis. And it had been increasing for a while. When we launched our system, we started to see trust in local government generally increase because the residents can see something that's being delivered to them that's very visible, that's fast. This is not an infrastructure project that takes years. For an elected official, they can decide to do this. Within weeks to a few months, you see the results on the ground.

And so success to me would look like Vias in every city in the US, and in Europe, and all countries that we operate in, and we're delivering meaningful value to local government across, of course, transit, but many other verticals as well.

Lance Glinn:

And so if we zoom out again, five, 10 years, look sort of big picture at the industry as well as at Via, what excites you most about, I'll say, the future of transportation and the future of innovation? And what role do you envision Via playing in the future of both?

Daniel Ramot:

So we can talk about air taxis and so forth, but I think if I were to just focus within the next five to 10 years, as far as surface transportation is considered, if we can have a system where you can show up in any city in the world and have access to an outstanding way to get around without having to rent a car, without having to own a car, simply through whether it's an app, or maybe there's a new way in which you access, glasses or whatever it is.

Lance Glinn:

Whatever new technology -

Daniel Ramot:

Whatever new technology will allow us to book something.

Lance Glinn:

Exactly.

Daniel Ramot:

We'll allow you to plan your journey and then pay for it and so forth. And we have a mix of these very efficient modes. They're buses, they're microtransit vehicles. If you have a disability, we can serve you very, very well. Whatever it is that you need to get you around where you need to go. And that is starting... Maybe this is a long-term vision. What I'm excited about is that starts to displace cars so people need fewer and fewer cars because this system is so great. I think that would be an outstanding vision. That would transform every city in the country into these, what I consider to be great cities, that have these great transit solutions, versus today where we're so dependent on our private car.

Lance Glinn:

Daniel, I've really enjoyed the conversations. I congratulate you on all the success that you, Oren, and the whole Via team have had. Thank you so much for joining us inside the ICE House

Daniel Ramot:

Thank you so much. Thanks for having me, Lance. I really enjoyed it.