Tony Zayas 0:02
Welcome to the tech founders show. It’s Tony Zayas joined by my co host Andy Halko. And here we’re on this show, we talk to people who are at the bleeding edge of technology, just you know, coming out with amazing innovations. You’re all about kind of where it came from, and the story behind it. So Andy, how are you doing today?
Andy Halko 0:23
I’m doing very well. How about yourself, Tony?
Tony Zayas 0:25
Doing good. Doing good.
Andy Halko 0:27
So tell us a little bit about who we’re talking to today.
Tony Zayas 0:31
Yeah. So today’s guest is Dr. Benjamin Schmidt. He is the CEO and co founder of ROADBOTICS. ROADBOTICS helps construct digital twins with visual data that’s meant for collaboration, engagement and evaluation, using the power of artificial intelligence, these interactive maps empower you to make data driven decisions about your environment, whether it’s in a large city, small community, or project site. So some super cool stuff here.
Andy Halko 1:00
My mind blowed by that description. So
Tony Zayas 1:06
how are you doing, man?
Benjamin Schmidt 1:08
I’m doing very well. How are you guys?
Tony Zayas 1:10
Real good. Thank you. So I obviously read the description, but would love to hear your take? How do you explain you know wha`t it is Roadblocks does?
Benjamin Schmidt 1:19
Oh, absolutely. So yeah, roadbotics is really helping, as you mentioned, sort of in this digital twin environment. You know, we started out working primarily with municipal governments, so things like smaller kind of local cities, local towns, things like that. And then also work a lot with like civil engineering, and construction firms. And one of the things that you run into when working in sort of this infrastructure environment is seems like a dub, but it’s massive, right? There’s tons and tons and tons of things, right? Bridges, roads, streetlights, traffic lights, right? There’s lots and lots of stuff. And unfortunately, kind of the state of the art right now is, you know, pen and paper. It’s jumping a truck is still around. And so like the idea of like, digitizing that information at scale, is still a very new concept. It’s where our company sort of works right now on that bleeding edge of, you know, how do we automatically identify automatically assess what’s going on in the world? And then use that information so that a government official civil engineer, can make better decisions about how to manage and maintain what they got. That’s how we’re going to get better infrastructures, better management, better maintenance.
Andy Halko 2:38
Wow. So and he just seemed larger story. Where do you start at to get here?
Benjamin Schmidt 2:48
I’m sorry, Andy, I think you’ve cut off for a little bit. What was the question?
Andy Halko 2:53
So I want to hear the origin story, how it started, and how did you get into this?
Benjamin Schmidt 3:00
Ah good question. So yeah, the company is about four and a half years old. We’re coming up on our five year anniversary, we’re actually spin out of Carnegie Mellon University. We’re based in Pittsburgh, whatever that means, this new, like digital first thing based loosely, in Pittsburgh. And really, it was one of my co founders. Actually, it was his original idea of, he started out with mounting a cell phone in a windshield of the car and driving around and collecting video data effectively in the first application, which we still do today. And our name, was looking at pavement distresses, like things like potholes and cracks. And since then, we sort of expanded the AI’s capability to look at signage, traffic lights, all sorts of different kinds of infrastructure. So that’s really where we started was on pavement surfaces, and then quickly brought in as we sort of realize there’s this much larger infrastructure world out there that sort of needs to help with this kind of technology.
Tony Zayas 4:05
And where did you guys like early on? Where did you identify that there was this market opportunity? For you know, what you had envisioned.
Benjamin Schmidt 4:14
I always feel like the coy answer is, you know, I’m a driver and I hit potholes. Yeah, they stop. That’s kind of the live version. I think we all sort of know, infrastructure has challenges. I think it’s the most common refrain, you know, I think when you used to go to parties, or like barbecues and stuff, you know, what a wonderful time that was. You know, I tell someone like hey, this is what I do for a living. You need to come to my town. You need my cities Rosa, the word right. I think every single person in the world if you ask them, you know what’s going on with your town? It’s the worst. We have the worst growth like okay, every everyone says that. It’s not necessarily true. It’s just hard to gauge but that’s certainly where it kind of it started is, you know, there’s a need for it. But I think definitively, you know, more on the sort of like startup space of kind of iterate all the lean methodology and VPs content, you know, certainly where my co founder, Christophe, you know, what his idea of using that smartphone, collecting video data, the first application was certainly on the roads, but we started to look at what are the other ones, right? Could we find and detect signs and sort of localize them within the world, we certainly focused heavily for our first few years on pavement, because that’s where we got the big response, right? Massive dollars go into management of roads, it’s a big budget item freaking, like a small town. And so focusing there sort of made the most sense in sort of chasing dollars and activity from a business perspective. But as we sort of, as I said, kind of advanced that technology, you know, the applications are really limitless and sort of understanding the built environment. And so we started to get smarter and smarter about like, you know, there are other kinds of stakeholders that might care about the water system, or, you know, utility poles and things. And so, from that we sort of branched out into other verticals, beyond just the government into these other spaces and sort of kept going. But that’s kind of, Yeah, the origin of how we got from, you know, potholes to where we are today.
Andy Halko 6:28
I’d love to dig into that a little bit more, you know, a lot in this show, a lot of people talk about minimum viable products. For you, you know, you talk about the cell phone and going around taking pictures, what did it start looking like out from there? Was it, you know, still a very manual process? Or were you able to get into technology quickly and scale it? You know, what did it look like, as you started to build it?
Benjamin Schmidt 6:52
Sure. So yes, we similar, we’re on the same train and lean methodology. So our first MVP, we have within like, a few months of our founding, you know, lots of development went into it. But basically, the first steps were moving from video, things like dash cam, and sort of processing that down into images and space, which is not really a trivial problem. But you know, pulling out individual frames every so many meters, so many feet, and then running those through an AI algorithm. So we had kind of two and we’ve continued to have really two parallel, like technical development tracks, right, you can almost think of them as like, separate MVPs that joined together until like the final offering, but for the client, one track is the actual machine vision, right? So here’s an image of, you know, a dash cam image looking out over the hood of a vehicle, what do you see in it? Like, that’s the first modern world, the other side of it is mapping. The other side is the mapping. Right? So okay, now that you’ve seen something, or you know, where the image came from, what what where is it? Right? Is it you know, 100 meters away? Is it five meters away? Can you localize that object within space, developing those two things was sort of the most critical, sort of baby stepping to ultimately a solution, which is, hey, I’m the Head of Public Works for municipal government, a small town, I have a half million dollar budget, where do I spend it? If you can solve those sort of two technical problems, you can then jump into solving the client’s problem, which is, this roads in good condition, this one’s in bad condition, this one, you know, might be teetering on the edge. And I could probably do a little bit there. And that’s really how we sort of finalize the MVP of we need to make the technical progress, progress mapping and AI, so that we could ultimately solve that client problem of like, where do I put my budget dollars to get the biggest bang for my buck?
Andy Halko 9:05
So what did it take to get over those technical hands? Like, did you need to bring in outside consultants is your are you and your co founder very technical? You know, for a layman, how do you then go from that space of like, we’re taking pictures that AI that’s helping make good decisions.
Benjamin Schmidt 9:26
So it was not outside consultants? It was it was all us. So we have been actually very fortunate. So this is not my first startup when the company I was at before. Actually several of the people there now work here, which made getting started incredibly fast. Right, like okay, well, you’re gonna work on that part, right? Yeah, I got this. Okay. So we did a lot of like the engineering work that way. You know, there’s a lot of iteration and we’re talking like four and a half years ago. So I think the state of machine learning deep learning now is even significantly better than it was then that we were buying, you know, Nvidia Titans at the time and trying to train algorithms. And then the other. Again, you know, one of the problems, I guess, or sort of challenge, let’s call it a challenge with doing things in the physical world is we spent a lot of time in the car, collecting data, and then coming back to the office and behind. That didn’t work I have, many of my colleagues will remember, we spent one Memorial Day weekend trying to collect an entire towns with a data, only to realize that my mistake, I forgot, I had a little I commented out the line that had the GPS being recorded. So we had tons of video, but we had no idea where it was, we had to go do it again. So lots of things.
Andy Halko 10:47
Which is common place for any startup, right?
Benjamin Schmidt 10:50
Absolutely. Come to the territory. Now, what what experiences what you get when you don’t get what you want? I got tons of experience.
Tony Zayas 11:00
Beyond that, Ben, I’d like to hear I think I saw, you know, in your LinkedIn profile, you have, you know, an interesting background, as you know, from researcher, and a scientist, CTO, you know, at the last startup you had, so, just love to hear a little bit about that backstory.
Benjamin Schmidt 11:19
I sure. So yes, I it’s been an interesting journey, I think, certainly chasing interesting problems is probably the most common theme. But yeah, I actually I have a PhD in bioengineering. So I did a lot of like brain imaging studies, as I said, the company before this one, but right after grad school was actually in the energy markets. So we were helping wind turbines to basically operate within power markets most optimally, but it’s sort of like predicting different components of what was happening. And then now here, I am in sort of a very infrastructure, civil engineering kind of market environment. So I would definitely say underlying all of it is this theme of sort of advanced technology and kind of interesting fields that I would hope can make an impact. I think that’s been maybe a theme here is I want to do something that really matters. There’s lots and lots of really cool technology. And to me, one of the most kind of interesting things, when you look across it is you can make incredibly fast strides in many industries by sort of borrowing from one’s you know, most advanced technology and applying it into a domain that maybe doesn’t know about that sort of innovation. And I’d like to think that’s, that’s really what we’re doing here. It’s sort of borrowing from a lot of other kind of like machine vision, localization mapping, things you’d see in a ROADBOTICS company, things you’d see in, you know, AVs, borrowing that same technology, and then applying it for helping governments make decisions help civil engineers to build and manage better. I think that certainly is the theme i i draw from my background is trying to find those kinds of opportunities to translate from one to the next,
Tony Zayas 13:05
I’d like to hear a little bit about some of the challenges that might have existed, you said, you know, you’re working in a space that a lot of pen and paper, right people do things in a very non technical fashion. And here you are with, you know, a solution that is AI machine learning, very cutting edge as far as what you’re doing, what were some of the roadblocks or hurdles to overcome to, you know, get in with government officials, civil engineers who aren’t used to this type of thing?
Benjamin Schmidt 13:39
A lot. You know, again, I would say that very much, there’s a lot of reception to using and sort of adopting newer innovations. I think everybody sort of wants that. The hard part is where the rubber meets the road. By the way, there are tons of road bumps. So it’s super fun, like roadblocks rubber meets the road. It’s great. Pretty fun industry. Ah, you know, the biggest one here is sort of almost, it comes down to trust. So you’re trying to create an algorithm. If you start almost at the beginning, you got pen and paper, why do you have pen and paper? It’s because you’re taking small amounts of notes over very large geographic thing. So I think about infrastructure for even a small town like a small suburban town go 45 minutes outside of any US city, any city in the world, you know, you have small towns, there’s lots and lots of them. But even they have a ton of infrastructure, like maybe 100 miles of roads, they have sidewalks that might be dozens or maybe 100 miles of sidewalk. How do you sort of keep track and manage all that? So pen and paper works as a way to record it? Now that the technological shift is like, well, you could certainly go and collect video data on all of that. That’s, that’s easy. Except them what you’re left with is hours and hours of video data. You know, gigabytes, maybe even terabytes of information on it, but no way to organize it, right. So like, the technology is there, but it can’t be unlocked. And to me, that’s the the AI sort of centric component of it is, you don’t actually want to look at all the infrastructure, that’s not helpful, but you want to do is focus on the places that you’re supposed to focus on, where there’s issues where things are arising that are unexpected. That’s what the AI can sort of do. And so the biggest challenge with that, you know, thinking about adopting our kind of technology, going from pen and paper to sort of what we’re talking about, it was somewhat understanding that kind of progression, right, that we’re going to help almost triage, right? It’s like, here’s where to focus, here’s not where to not focus. It’s also then trusting that that triage process works. Now, again, one of the huge benefits that we get is that it’s very easy to verify, right? So it’s almost always feel like it’s a kind of one of those NP complete problems, right? Solving the triage problem is super hard. But it’s very easy to verify, because you can just go look at all the data and say, like, oh, yeah, you are correct. So having that sort of dual mode was always something that we found very helpful, which is, we’ll give you all of that video data that you can’t look through all of it. And we’ll focus areas on on where you need to pay attention. And if you ever wish to actually check, you can go look through all of it verified, right? So like, we can stand by the promise, and that makes that trust barrier, lower significantly, which makes it easier to adopt and move forward. So that, you know, I think trust is always a, for any innovation is the biggest sort of question, Can you can you get the new process to work is really, can you get people to trust the new process works better?
Andy Halko 16:58
And how did you get your first couple of clients on? What was the the approach to having them trust something that’s brand new?
Benjamin Schmidt 17:10
Lots of meetings. There’s definitely lots of meetings, lots of demos, lots of iterations. And that’s probably the shortest answer. The reality is, you know, what, we spent a lot of time in kind of Western Pennsylvania driving from one town after another to sort of showcase what it is that we did, I think very much in the bringing up Lean Startup, if you want to do the Crossing the Chasm kind of thing, there are definitely the early adopters. And every market government is included, there are the people who want to jump on that bandwagon. And so finding those is kind of the first search problem that you have, identify those, showcase it, prove it, they’ll adopt it, and then hopefully become your sort of use cases for the next ones. But I think that’s that’s really what it comes down to, it was just kind of boots on the ground, making those kinds of things iterating showing that, you know, feedback is accepted. So there would be, especially at our early days, it’d be one of our kind of now infamous for screw ups was that our algorithm was not very good at or we had done most of our training in, in the summer, I think it was, and then come fall, we started collecting new data, but there were leaves all over the ground. And the algorithm started thinking those were cracks and bad. And so started reading them poorly. That was a whole sort of like, gosh, alright, now we have to go back to the drawing board. You know, the client points out that, like, what are you doing here, like this road is fine. You know, and trying to create that sort of virtuous cycle of both trust but also iteration on our product development to make sure that we’re solving those problems. So all of that it’s really, you know, we didn’t do it once and we were finished. It was try Okay, try again. Okay, try again. Okay, try again, years later, we can now say that, like, we really do have it, we have our process, we have our system, verified by lots and lots and lots of mistakes.
Andy Halko 19:12
And I will share I grew up in Pittsburgh, most of my life. So I am a Pittsburgh boy at heart. I would love to talk about AI because we this is a tech show. And we really like to dig into the tech. You know, where is it right now? And where is it going from your perspective?
Benjamin Schmidt 19:33
Well, that’s a very big question is, so I can you know what, I guess again, we have sort of two parallel tracks of how we sort of think about AI. I don’t know if this is sort of the normal flavor when it’s how we always think about it. We have the machine vision part. Here’s an image extract, you know, put bounding boxes on it, semantic segmentations, etc, pull the things that are meaningful at the image. And then we have the other side, which is that localization problem So, you know, not usually in a single photo, but like in video, we can do some of the simultaneous localization and mapping and try to auto find, where is the camera perspective inside there? I would definitely say on both dimensions, you know, thinking about our own sort of r&d efforts, the big entities continue to make huge strides and then open source them happily, that benefits kind of all of us smaller companies that can’t feel a $100 million research and development team. So I think that’s a huge source of it. So in some ways, we’re sort of beholden to them on on what their most their latest model development is. That said, you know, that’s kind of like, oh, no, you know, what is Google and Facebook doing? And, you know, will they one day show off the tap? I’m less worried about that problem when it comes to sort of AI overall, the one that’s way more interesting to me is, application is still incredibly, incredibly hard to get right. So you can have like, state of the art AI’s, but actually getting the application right is the hardest part, how do you deliver it? How do you distill whatever’s coming out in the algorithm into something that the client cares about? My belief is, you’re more likely to die as a startup on that application problem, then you are worrying about what the next
Andy Halko 21:34
no big deal. Yeah, you know, AI is very interesting to me of, I keep wanting to talk to people about where it’s headed, and how is it going to change the world, because we are making these very quick strides that are just changing things really fast. And so, you know, for someone like you that’s like really tapped into it. I mean, you you kind of have a viewpoint that I think most people that are either touching tech a little bit, don’t have or especially for people that don’t know anything about it.
Benjamin Schmidt 22:10
Absolutely. Well, so I think maybe an answer in two parts. I think in the long arc, I totally agree with you, I think AI has some massive potential to sort of change and shake up things. Change the way that we do business, all that kind of stuff. It’s really, it’s, it’s not the cool, exciting answer that you sort of read about all the time. But really, most of the challenge most of I think, I would imagine, you know, in talking to like other founders, most of the real challenge here is not getting your AI, right, that’s a, that’s the necessary but not sufficient condition. Right. You can do all that you can make some cool AI, you can have it do all sorts of things. But if you cannot find a product, you can’t find someone that buys it, it’s It’s all useless, right? We used to have a joke that like, you know, we could build a model that sort of predicts how many sodas are in the fridge right now. But who cares? Like, you could be the most advanced model in the world, you could run sort of massive deep nets, you could train it for weeks, you could have it on 1000s of CPUs, it could be state of the art, who cares? It’ll have no meaningful impact on the world. And I think to me that finding that that point in between is what’s so hard as entrepreneurs is, you know, we get sucked into the the more the excitement, the like, oh my gosh, that. Then next data set, they pass this they solved this problem, like incredible. Who’s gonna buy it? How am I going to get it out into the market? Can I actually put it in a form that’s digestible? And does it solve anyone’s actual problem like that leap is still so huge that, you know, again, I think that’s where we’re gonna, we’re all gonna live and die is on sort of translating it. And is often to the theoretical but you know, I think about reverse, okay, imagine tomorrow, they come out with an AI that has human level intellect, terrific. I still have the same problem. With the great I have human beings that work at ROADBOTICS, I’m a human being we work at ROADBOTICS, our problem is to figure out what our clients want, right? So even if we have aI level, human level AI, you still need to find something that people want and then deliver it in a way that makes sense. So I’m super bullish on sort of what the prospects are, I think some of the newer deep learning tech even some of these like reinforcement concepts that are starting to become more and more popular, like the playing Atari and all that kind of stuff. Very cool. If you think like, take our industry, how do you apply it? How do you make that worthwhile? How do you help governments? How do you help me to spell these? That’s, that’s the main game.
Tony Zayas 25:04
So just to dig into that a little bit, how did you and ROADBOTICS, how did you guys go about the process of addressing that application problem? And how do you make that, you know, compelling case to that target audience so you can translate that value of what you guys could do? Allow them to see, you know, what that would mean to them? Because, again, it’s so different than what they’ve been doing. How did you guys go about that process?
Benjamin Schmidt 25:36
Ah, well, we certainly didn’t get it right. At first, I’d argue we might not still have it right yet, but we’re gonna get closer and closer. Very much iterative. I think everything, everything we’re talking about. It’s all iteration. Right? We try one to see how it goes. I think the disingenuous version will be, you know, we’re throwing spaghetti at the wall, see what sticks. I’d like to think we’re a little bit more advanced than that. But, you know, on my pessimistic days, I don’t think we’re that much more advanced than that. It’s just how fast you can throw spaghetti at the wall. So that I think is very much how to find that iteration loop that kind of creates the work, you know, you get feedback from the client, that’s the most important thing. If you don’t have any feedback, it’s useless, right? If you’re just throwing product over the wall, and then not paying attention. You can’t iterate. So getting feedback, that’s a huge part of it. I mean, we have a whole team that’s just dedicated to that, which is, what are the customers think, how are they using it? What do they like? What don’t they like? What do they ever press that button, that kind of stuff? So that’s a big part in kind of finding that product market fit whatever you want to call it? So yeah, I think iteration, it’s all about iteration.
Tony Zayas 26:51
So just to learn a little bit more about, you know, your co founder as well, you guys are both technical founders, correct?
Benjamin Schmidt 27:00
Yes, yeah.
Tony Zayas 27:01
Did you find that, you know, part of the challenge was, you guys are very tactical, you’re into this solution into the weeds developing this thing out, was part of it, you know, how do we convey that message to that audience create something that really resonates with them?
Benjamin Schmidt 27:22
I’m not arrogant enough to think that I would say that. No, that was easy. The part I would say is that, so our previous company Conterra. I think probably the lesson we took away, especially when it comes to sort of selling advanced technology is don’t sell advanced technology. Wow. And yes, I have a technical background, I’m very technical by nature. It pains me to think that this is actually how the world operates sometimes, but it is definitely how the world operates. No one wants to buy your AI solution. Nobody, nobody’s gonna be like, excellent. It has AI, I will purchase it. Right? Like maybe if you maybe some, you know, really crazy tech people. Or if you’re directly selling AI, that’s what you’re selling caveats there. But for most, we do not walk into a government and say, like, hello, you should buy us because we have AI, right? Like, that’s not usually buys, because we solve your problem. That’s it, you know, how do we solve it, we use AI, we use a whole bunch of technology, that’s what makes it bigger, better, faster, that’s the why. You’re not going to buy it because of the why you’re gonna buy it because of what we can do. And I think to me, that’s one of the lessons that definitely has taken a long time to learn. But certainly at ROADBOTICS, we applied it, I think reasonably well, we always get caught in the trap of once in a while, but we use AI to solve problems. But the problem we’re trying to solve is you do your client need to manage your infrastructure, and we make that easier. Now, if you want, I can talk to you all about how we do it. But for most people, you know, for most sort of like heads of organizations before certifying authority, whatever you want to call it, sort of your sales organizations. That’s what they want to know. Are you gonna save me money? Are you gonna make me money? You know, and can I rely on you to do the two things, you know, one of those two things you said? Great, terrific. So focus there. That’s that’s the the message. So? Yeah, I think even for ROADBOTICS, I think, yeah, we have some really cool technology. We’ve done some amazing things with that. But when we are sitting down with the government, we are talking about potholes and fatigue cracking and signage, inventories and you know, can we get water reports going and how are we going to talk to council and unlock grant funding? That’s what they carry. Right? So that’s what we focus on. So that’s, that is definitely, at least the lesson we’ve taken away from it is we’re not selling technology. Were selling the solution to your problem.
Andy Halko 30:07
You talked about the application side of it being really a challenge. So how did you face that of taking information that you created and then building away to really deliver it to your client base? You know, how did you look at that from of trying to create a solution from that perspective?
Benjamin Schmidt 30:28
It’s, it’s a good one. I think I think it’s a misattributed quote, but I like it. Anyway, the If Henry Ford had asked people what they wanted, he would have said faster, or they would have said faster origin. You know, I think very much the the hardest part was trying to find those things is that, you know, we talked to a lot of clients, we’ve continued to talk to a lot of clients. But no one is ever going to tell you like, this is exactly what it is that I mean, they’re going to get close, they’re going to tell you like, where their pain points are, they probably do it indirectly. But human beings just sort of suck at telling you like this, if you build this exact thing, this is exactly out like now that’s, that’s our job. That translation is, is probably the hardest part. You know, and we’ve gotten burned by it before we’ve been successful with it before. But yeah, the like, if you give me this, I will buy this, you know, that should always set off alarm bells for every entrepreneur, like, I don’t think that’s true. I really don’t think that’s true. I think that’s what you know, you want to happen. Yeah, I don’t know that. That’s really what’s gonna, you know, if I put if I put this in front of you, ah, this solved the problem. Trying to find that is really, I think, that’s the whole problem as entrepreneurs is that search problem. Here’s what the feedback that I have. I can’t 100% say like, that’s exactly what I need to do. That’s not my roadmap, right? Like, you have to have your own sort of view. One of the things that just because of who we are. So I’ll call it a benefit, but it could certainly be a detriment. We are not civil engineers, right? I’m not assuming. So we’re working in a space that we work a lot with civil engineers, public works officials. But we don’t come from that space, which does give us I think, sort of a unique perspective on it. And that we are trying to learn the ins, especially the beginning, you’re trying to learn the industry, while also trying to identify, what should we do to help and go back and forth. I’ve talked to other people about this sort of like which way it works, but you have this sort of, I’ve been in this industry for 10 years, and I know exactly what we can do to improve it. And then you have kind of our version of it, which is we’ve never been in this industry. And we’re looking at it from an outside perspective, knowing a whole bunch about like, what’s possible in other industries. You know, I don’t know, which is better, but certainly the path that we’ve been on is that second one, which is okay, so wait, how do you do that? Like, oh, that’s what you mean, when you say, okay, like, here’s the new vocab you’re talking about. And then that starts to get our own sort of, like, iteration on like, you know, actually over in this industry, this is how they like, what if I gave you this? That was all it like. So that that both unique take and again, iteration, I think is, is very helpful in trying to find that right answer.
Andy Halko 33:30
Yeah. So you’ve since you’ve done multiple startups or been involved in them, you know, what’s your viewpoint on bootstrapping versus raising capital?
Benjamin Schmidt 33:42
Oh, man, that’s it. That seems like a setup. Well, uh, well, so I mean, for us we did we did bootstrap for the first like, eight months, which came with quite a lot of things. So we did that. And then we did get our first fundraising. We are now today venture backed. I, you know, ultimately, I think it’s just it depends on sort of how big your market is what your capital costs all that usual, like business pieces of it. But at the end of the day, we phrase this correctly, because I’ve run into this a lot, I think, people who are inclined for bootstrapping, and then look at sort of the the venture capital model come in with this view that like venture capitalists are going to force them to do all this, like, on a day to day basis. That is just not true. Venture capitalists are also people they want all of us to be successful. They know we’re human beings. They’re like, we’re all trying to build something and accelerate it quickly. Yes, their business model is assuming that like one in 10 work, etc, etc. But they’re not cold and callous about it right? Like it’s just that’s how it is right? We’re going to try to scale this thing. I think the biggest one is just going in eyes wide open what it is that you’re trying to accomplish, right if you if you want to several million dollars and staff up and try to hire and go on this like massive growth trajectory, that’s venture capital. You know, unless you have a lot of money. You’re not bootstrapping that kind of company. The flip of that is, yeah, you can, you know, bootstrapping does work. You can grow companies that way. There’s nothing to sort of preclude it. But I don’t think necessarily, like one is better than the other. It’s really just a matter of like, what’s your, what’s your trajectory look like? Right? Are you going for rocket ship? You know, and it’s a rocket that either goes to space, or you’re dead? There’s no in between? You know, there’s no exit? Or are you trying to do for the slow and steady, but you might never get to space, but you at least be airborne? I don’t know. That’s, I think, for everyone to judge for themselves. But, you know, pros and cons on each dimension. I don’t think it’s necessary. Like,
Andy Halko 35:52
Yeah, I agree. I think we’ve right, and we’ve had founders on on both sides of the coin. But I think you’re right. I mean, it’s always a very specific case. But there’s a lot of founders that you always hear sitting there on the fence and saying, you know, how do I approach this? Do I want to try and just like grind it out? Or do I want to, you know, spend my my full life trying to pitch folks to give me money. So,
Benjamin Schmidt 36:17
exactly. But uh, yeah, and this is the other I mean, certainly on my last comment, this is another one that I hear a lot of like, and then you have to try to raise again, it’s like you’re always raising whether you’re bootstrapping or not. It’s just whether your customers are the ones funding it or whether the investors are the one funding it. We never stopped selling I feel can just get in quotes all the time. But one of my other favorite ones is the hierarchy go in an organization, the more every role looks like sales, I can definitively tell you that it’s 100%. True. Like, absolutely. Every single day is selling. It could be a venture capitalist, it could be a partnership, it could be a client, you’re always selling, it just depends on which one which thing? Could it be product could be the company itself. Always selling?
Andy Halko 37:00
I love that. Yeah.
Tony Zayas 37:03
So just to shift gears a little bit, what is the team at ROADBOTICS look like?
Benjamin Schmidt 37:10
A good question. I think we’re at like about 30 employees right now. And yeah, I think we’ve got the usual sort of grouping of individuals you’d see at a startup. So we have a sales team, we’ve got marketing, our engineering teams, kind of a r&d group. But the one that’s a little bit unique given some, what we do is, is really, we have a decently sized services department. So we’re in kind of this, I don’t know what it is, but like a limbo ish state where, you know, we use a lot of AI, but we absolutely have to be right. So we do a lot of like, quality control. This is something we learned very early in our process, about, you know, machines do most of the work, but we still need someone to make sure that it doesn’t do anything crazy. The reason being that, you know, as I said, hundreds of 1000s of dollars were being sort of decided upon what we’re doing. So being wrong has really bad consequences. But if that’s kind of what you meant, I mean, that’s the usual sort of break up or break down here, sort of how we’ve positioned ourselves. It’s ebbed and flowed on, like headcount in each department and sort of, you know, some grow faster than others, but no, get I think we have an outstanding team. It’s one of the reasons that we’ve been successful.
Tony Zayas 38:32
Yeah. And along those lines, do you have any, you know, any advice that you have for hiring? How do you find the right people for your organization? And what would be, you know, just a nugget, just that you’ve learned over the years that you would pass along as a little bit of advice about finding good people finding the right fit.
Benjamin Schmidt 38:56
A nugget of advice, I have no idea. Hiring is hard. Hiring is the hardest thing ever. I mean, all of your success will come from the people that you pick. So again, I think like taking lessons as a technical founder, right? Your product, your company, your technology, they’re largely irrelevant if you have bad people, or at least not, not the right people, not safe people or bad or something like that. Just not the right fit. It will really come down to people. It’s all it’s all people. So I think when it comes to hiring lots of different philosophies, we’ve made tons of mistakes. We’ve made tons of successes. I think some of it is really that. I think others talk about it’s like a gut feel. I think just over time you get enough experience with it and you start to feel like this is going to work this is not but you never really know. And you know the other part here is I think we’ve all experienced this over the last, like 18 months is, people are people, they’re complicated. Other things can happen and life events can happen. You’re never really quite sure of what happens next. But I think that’s just part of the journey that all startups go on is hiring people retaining people, keeping it exciting. And you know, very much I’d like to think, you know, try not to burn bridges and things like that. You never know, especially, you know, for us, like in Pittsburgh area, small community, and we all do we move from startup to startup, so you can see all these folks.
Andy Halko 40:39
How about your role in the company, you know, for a founder and founders that might watch this? How, how has your role changed from like, day one to now? And how do you see that evolution of a founder over time?
Benjamin Schmidt 40:56
Um, yeah, man. I don’t know. I’ve take like, almost five years to explain, I think. I don’t know. It always feels like a real music quote. Yeah, there you go. I’m all out of my freshman. Ah, you know, I think how has it changed? You know, certainly it’s in the beginning, your every roll. Towards the end, it’s more like you know, hopscotching around from role to role. I think that’s, you know, at least the one I can most imagine, I think a lot of the, the, like, internal growing pains that we’ve gone through is like, you know, okay, this job is, you know, this person, and they have multiple jobs, and then you start to like, sever those and break them apart. Like, now this person is responsible for this. Because they’re each now too big. I think those are a lot of likes day to day, or as things progressed, jumping from every problem to like, Okay, now. Now we have people who are experts at that, like, great. I’m not, I’m gonna go over here. That’s super helpful. Always love that. Hire for that. That’s always a good idea. I think that’s probably the biggest change. But I would definitely say the the roller coaster always exists, you’re always going up and down at one moment on the top of the world, and everything’s winning. And then the next one, when it’s like, oh, my gosh, it’s all crashed. The differences. Yeah. As the organization gets bigger, it’s like, the roller coaster is like, the whole company is out. Woohoo, we’re all excited. And then the whole company is down under small. When you’re big. It’s like, this team is really excited. And then you look at that team. It’s like, Oh, crap, everything scratch. So it’s always the roller coaster. Maybe the frequency, or the amplitude changes, but the frequency doesn’t seem to change.
Andy Halko 42:49
I love the analogy, though. I mean, I agree with you. I’ve been doing my business for a long time. It’s definitely a roller coaster. And, you know, as you get bigger, it kind of just, it changes. But, you know, I’ve always felt like good entrepreneurs can make it through that roller coaster. They’re kind of staying somewhere in the middle. Even though everything is going up and down that you know, mentally and kind of emotionally, they can at least somewhat stay in that middle space and don’t go crazy.
Benjamin Schmidt 43:19
Oh, 100%. I mean, I like that idea of like stand in the middle of space because it very much. Especially at the early days, you feel like every single one you’re like one misstep away from it’s always a fun thing, which I knock on true. I do think the biggest one here, it’s like it’s going to happen tomorrow. It’s also going to be a roller coaster. And next week, it’s going to be a roller coaster. I think you mean things like vacations. Now. Okay, yeah. Next week, do you want to take off like, the roller coaster is going to be there when you come back. It’s going to exist, whether you’re there or not, like, I’d love to think I have an effect on it. But it’s not going away. Like you’re not gonna, I’ll just do this one last piece, and then it will be solved. It’s like, No, this piece will be solved for a small amount of time. But that piece over there that you weren’t paying attention to why solve this one. Oops. But yeah, I think like you said, level setting is really important. It does get easier as organizations scale. Not to mention, like I’ve said, I mean, there are 30 other people at the company who have way more expertise in their areas than I do. It’s like, great. I could sleep better at night knowing that that’s what they’re doing and I’m doing this part. Terrific.
Andy Halko 44:36
Someone else can firefight that fire.
Benjamin Schmidt 44:39
Exactly. Yeah, much smaller. You You are definitely fighting all of them.
Andy Halko 44:44
No, I agree. How have you, certain way, that you’ve, I guess manage people through that roller coaster. I always when I talk to founders, it’s like they’re going through the ups and downs but that means their team is to and You know, of course, we want to hire people in startups that like a little bit of the roller coaster because they don’t want the day to day. But still, you got to somewhat manage your team through the craziness as well. Have you had to do that or any insights on how you do that?
Benjamin Schmidt 45:17
I do think well, maybe to flip the question a little, what I think is really kind of fascinating. And again, we’re not that big of a company, right. But as we sort of mature and grown, at the very early days, I think, almost every single person, and it’s really like, go through the list, but like, I think every single person for the first like, at least few years, you’d ask them like, hey, what do you know? Don’t ask him a boring question, what what’s your five year plan? But like, if you asked him something along that line, right? It’s like, well, I want to do my own startup or I love start, like, it’s very much like they want that experience exposure, they’re interested in sort of the journey part of it, right, they’re in it for somewhat the rollercoaster, the, and they’re like learning how to adapt to it, if they don’t already have experience in it up, like, this is what it’s like, every single day. And for the most part, you know, they’re thrilled about it, right? It’s always kind of exciting to try that for the first time. The, as you said, level setting the endurance problem of like, you know, try to do this for 10 years. I think as you sort of like step through that maturity progression, you start moving from kind of that of, you know, I want to do startups to more the professional, you know, this is what I do in life, like, this is my role, right? Like, the startup thing is just kind of, oh, yeah, that’s neat, you know, a little small edge back here, but like, it’s not. What do you want to do if we made a big exit? Like, I’m gonna start my own cup? Like, that’s not the answer anymore? Which is great, right? Because that means you’re starting to get kind of diversity of opinion around around the table as you’re like, Oh, well, you’ve been doing this role for 10 years, you definitely know what, how to answer this question. I think that managing question that you’re asking, really evolves in each of those phases around sort of how to manage that roller coaster. Just different kinds of people and each one.
Andy Halko 47:20
And just the kind of follow up on the roller coaster pieces with COVID. How did that affect you? And then more so like, how did you handle this big outside force that kind of came in and upset everything for everybody?
Benjamin Schmidt 47:38
Yeah, that was was fun. That was definitely tough. I think it was tough for everyone. Because I think it was just massive, massive uncertainty. I still remember we had come up with if the governor close the Pennsylvania or No, no, it was if the city closed their schools, we close our office. And I left on a Friday afternoon. And then yeah, that was and then Governor Wolf, Pennsylvania actually closed all schools across the state. And I was like, Okay, everyone got on Slack immediately. It was like, Ah, I guess we’re not coming to the office on Monday. Happily, we’d have a little bit of time to sort of prep for remote work, but yeah, not easy. Certainly not something I hope anyone else ever asked to, like, you know, take a lesson from this and do again in a few years, like hopefully, it’s just that’s not a thing. But I do think the kind of adversity, scrambling, etc. Those concepts are roughly the same. This one just had that particular character actually, potentially harming everyone involved, that that part was pretty devastating. But you know, I think when we do our jobs, right, kind of as entrepreneurs or something that we want to call that people are ready for the craziness, right, the most hectic things you can imagine the scrambling this sort of like getting ready. You know, we don’t have a lot of rigid systems. That’s probably a con in some ways, but it’s also a big pro when it comes to perturbations and things that really shake us up. So yeah, not fun. But I’m glad that we’re sort of still holding on and we’re able to kind of get through on the other side, or hopefully on the other side, without too much sort of tumult.
Tony Zayas 49:32
So then I would like to get just your take on where do you see the business going in the next 12 years? Kind of what’s on the horizon? And then what’s the three year plan path? What does that look like?
Benjamin Schmidt 49:46
I like it, did it 12 years initially?
Tony Zayas 49:48
No, I said 12 months, okay.
Andy Halko 49:52
12 years too. I was like, man, Tony, you are really this is gonna take a lot of imagination. Okay. I’m working bad
Benjamin Schmidt 50:06
12 months kind of three years. I mean, it seems boring, but I think growth will be started. Certainly a big part of it. Yeah, I think we, as I said, we’re definitely on the venture capital roller coaster, or rocket ship, whatever you want to call it. So my guess is, you know, I think all expectations are somewhere by the end of next year, we’ll be looking for sort of like our next raise, that’ll put us into a totally new sort of stratosphere of capability development capability, right, we can really start to supplement a lot of our product development. What I think is the most exciting is that, while you look at kind of like, what’s happening in Washington was very specific to our business. All the infrastructure discussion, you know, I think there’s the funding part of it. But then there’s just the fact that like, when you talk to anybody, and they know about, like, oh, yeah, infrastructure, yeah, infrastructure, like it’s becoming something that we’re all paying attention to. And I think is creating this really great tailwind within the infrastructure construction, you know, this built world kind of environment, then I think for the next several years, we’re going to be able to really ride that into some bigger success into certain sales numbers, etc. But like, I think we can make some massive strides on on product development, simplifying how managing infrastructure works. And ultimately, I think the big takeaway would be, no, it is painful to know that, you know, this is the world we work in, and yet it’s like, you know, America’s infrastructure is crumbling, right. That’s the usual refrain. We’re gonna solve that by doing better maintenance and management. And and we’re not going to do that by just kind of like throwing money at it and hoping it happens, right? It will be technology that figures out how to do that bigger, better, faster. So I’m really excited that what we’re building what we’ve been building is just kind of like, ideally suited for that problem. So it’ll be on us to make sure that we can weed that path and, and really take advantage of what’s what the opportunity looks like. So I’m excited. I think, really, we have a chance to make a big, big impact here.
Tony Zayas 52:23
Yeah, now that is exciting. And a good point that you guys probably will have that tailwind because of the spotlight on infrastructure. And yeah, exciting times. Yep. Very cool.
Andy Halko 52:36
So one thing we didn’t talk about, just real quickly, are there a lot of competitors in this space? And how do you look at competition? Do you pay a lot of tension and worry about it? Or just say, you know, this is our path, and we’re gonna stay focused on it?
Benjamin Schmidt 52:50
Ah, well, there are sort of like a few other kind of adjacent to us, I think there are some doing kind of close or close-ish, but not sort of direct to to competitors. But I think to your, your kind of last part of the question, don’t necessarily look like I think, can come back to that idea of like, we just need to solve a problem. You know, there are a lot of problems out there, they might solve one version of it, we might tell the other and that’s fine. And my philosophy is always we are going to determine whether or not we win, we’re going to determine whether we get the product market fit, whether we can get the right sales proposition, it’s not going to be some competitor company to our launch, unlikely we’re basically going to win or lose our own ability to execute.
Tony Zayas 53:46
Well, then, before we dive into probably our last question here in a second, tell us where you know, for the viewers where they can learn more about ROADBOTICS, follow what you guys are doing, see what you’re up to?
Benjamin Schmidt 54:01
Absolutely. So I’m very active on like social media, or ROADBOTICS as accounts and things. You know, we talk a lot about kind of our research and the new technology, but if there are any people out there that are in that civil engineering in the government, space, infrastructure, ROADBOTICS.com that’s the easiest way to find out about us, but everybody knows somebody in government, that’s usually how it works. And no point of our way. We’re gonna help.
Andy Halko 54:31
Awesome. So just a final question that I asked everybody. You know, I’m really interested in if you were able to go back in time to before you started the company and have coffee with yourself, what advice would you give,
Benjamin Schmidt 54:45
Oh man? How many how many minutes do I have to tell myself about the future?
Andy Halko 54:50
Right. Back?
Benjamin Schmidt 54:55
Yeah, um, well, the boring answers would be the things that we discovered are the right product market fit But that feels almost like cheating. I would probably actually, you know, if I could go back in time, I would tell myself to take it slow. Less chaos, just like what you’re talking about kind of the level setting. That to me is the absolute biggest one. I think it’s a probably the best lesson that I’ve learned over the last several years here at ROADBOTICS is, yeah, take it slow. Tomorrow, it’s gonna be chaos, too. But the worst thing you can do is make stupid decisions. Just keep on trackin.
Andy Halko 55:38
Awesome, thank you.
Tony Zayas 55:40
Yeah, we’re very good. Well, everybody. Thanks again. Dr. Benjamin Schmidt from ROADBOTICS. Thank you for your time, man. We really appreciate it to our viewers. We’ll be back again next time next week with another exciting interesting founder doing amazing things. And we will see you then. Take care everybody.
Andy Halko 56:02
Say Thanks, Ben.