From Sidewalks to Scale: What It Actually Takes to Build Real-World Robotics | The Pair Program Ep96

Welcome to The Pair Program from Hatchpad, the podcast that gives you a front row seat to candid conversations with tech leaders from the startup world. I'm your host, Tim Winkler, the creator of Hatchpad. Join us each episode as we bring together two guests to dissect topics at the intersection of technology, startups, and career growth. Welcome back to The Pair Program. I'm your host, Tim Winkler, joined by my co-host, Sean Leahy. Uh, Sean, I got a, a scenario for you. All right. I want you to build your ideal delivery stack. So you got 3 options. You've got an Uber driver, you've got drone delivery, and you got a sidewalk robot. Uh, who are you gonna use for speed, reliability, and, uh, a late night, we'll call it, you know, maybe a little drunk munchies order. Interesting. All right, so speed, reliability, and then we'll say, um, reprovisioning, um, for the kids out there. Yeah, for our sober-minded listeners, I'm definitely not going with Uber for reliability. Um, for reliability, I might actually be tempted— and I, I'm not just plugging our guest here— but to use one of Cocoa Robotics' units. Because it's got that advanced computer vision on it. It's good for speed. Yeah, for speed, I, uh, you can't go wrong with, with, you know, like an MQ-9, you know, some of America's finest drones. Or, uh, I should have, should have mentioned Anduril. Oh well. But yeah, I'll put the drone for, for reliability, for speed, sorry. And then for reprovisioning, you know, that's when you get into an Uber and you make some friends too, right? If you just, you hop in there when they show up 25 minutes late but you're going to McDonald's anyways. So, you know, whatever, it's a little adventure. But that's, uh, I don't know, Tim, what do you think? What's your forecast? If you're going to invest in the Sean Delivery stack, do you think this is going to be a viable enterprise? I mean, I dig. I think you had valid, you know, points for each, uh, kind of platform. It's tricky for me. Um, I think a lot of this is location-based, right? You know, I'm out a little bit further out in the burbs. Um, I don't know how far the sidewalk robot's going to realistically travel to get me that bag of spicy sweet chili Doritos, which is the elite flavor. So I think for my current situation, I might lean Uber driver. But if I'm putting myself in more of a dense city, let's call it like it's morning rush hour and I need that breakfast burrito, I'm thinking I'm leaning robot on that one. Not really trusting like an Uber to drive, you know, through traffic, get there on time. And then a drone just feels like it's going to get pretty, pretty messy with a burrito traveling at high speeds in one piece. So, you know, I think, I think again, you know, it's a little bit location-based and, but I get why people are starting to lean into the robot side for delivery. And that's where things I think get interesting. We keep hearing about you know, drone delivery and autonomous systems for years, uh, but I— it feels like few of these technologies actually show up consistently in the real world. And, uh, that's really what today's conversation's about. You know, it's not about just building robots, it's about getting them to actually do like that work in messy, unpredictable, real-world environments and, and doing it in a way that scales. That's the hardest part, in my opinion. It's not just the tech, it's, it's everything that happens once you leave the lab and get into those real-world settings. And so we're going to be unpacking that today, what it takes to not just build robotic systems that operate in the real world, but turning that into a business that's actually scalable and defensible. And so we've got a couple of great guests joining us to unpack this. First up is Zach Rash, co-founder and CEO of Cocoa Robotics. Zach and his team are building one of the largest autonomous on-demand delivery fleets operating in real cities today, powering hundreds of thousands of zero-emission deliveries across the US and Europe. Zach, thanks for joining us on the podcast. Yeah, thanks for having me. Yes, sir. And then alongside Zach, we've got Paige Craig, managing partner at Outlander VC. Paige was one of the earlier investors in Cocoa, and he's going to bring a a very unique lens around what actually makes companies in robotics, uh, and more broadly, broadly, I'd say like physical AI, you know, uh, what makes them defensible. Uh, and so it's what I love about, you know, the Pair Program format. We've got two perspectives kind of tackling the same problem here, one from the operator and building side, and then one from the investor side, thinking about, you know, call it like these long-term moats, uh, in execution, uh, for, for robotics and delivery. So thank you guys both for spending the time with us. Uh, we're excited to to dive into this. But before we do, we always kick things off with a quick warmup. We call it Pair Me Up. We'll go around the room, we'll rattle off a couple of things that go well together. Sean, why don't you lead us off? Sure. So Tim, today my pair is spring and spring cleaning, right? So we're finally at that stage of Virginia weather where it's probably not going to get down to freezing again. So I'm going through all my stuff and finding things I can throw out. And I, I think I'm at that age now where I just want to get rid of most of my stuff, right? You spend a lot of time accumulating and then you realize you have too much. And so now I'm like filling up my F-150 week after week with things to take to the dump. So, uh, for everyone out there who has too much stuff in your house for spring cleaning, or, uh, to manage, spring cleaning is a good thing to pair maybe with that bag of Chili Lime— was it Chili Sriracha Doritos? Yeah, Sweet Chili. Sweet chili. Yeah. Yeah, man. I think it's obviously super relatable. Spring cleaning is kind of one of the things that I love and dread at the same time. I know it's going to be a heavy lift, but then when you get it done, you kind of just feel lighter. You feel good about getting rid of the stuff that you've accumulated over the winter. You know, a lot of the stuff from the holidays, we're actually getting into it this week. So this weekend, so I'm We'll make sure my wife knows that you mentioned that, Sean. We'll put her off on a, on a good step today. I don't think we've ever actually done any spring cleaning before now that I think of it. Oh, sorry. You tell this to my girlfriend. We're probably due for some of that. Yeah. We'll have, we'll have her listen to the episode, get her excited for it. All right, cool, man. I'll, I'll jump in. I'm going to go with newborns and nurseries. Uh, so I'll go ahead and make the formal announcement here on the podcast. My wife and I are expecting our second child. Congrats. Thank you. Outstanding. Best news in the world. We've got a baby boy coming this summer. Uh, we've got a 3-year-old daughter, Alice, uh, already. And so we're excited to give her a baby brother. And with that comes, you know, a bit of the reshuffling at home. Uh, so dad's getting kicked out of this beautiful home office I'm sitting in. Most of our listeners have probably gotten used to seeing this on our YouTube, but I'll be relocating down to the basement. We'll convert this to the nursery, and there's a real chance that come June or July, I'm going to be running episodes with very little sleep and maybe feeding the baby in between segments, and might find me in this backdrop turned into a full jungle theme at that point. So I'm leaning into it. Newborns and nurseries. That's, uh, that's my pairing for, uh, for this one. Well, Tim, definitely want to give a hearty congratulations to you. Like I said, best news in the world. And, uh, it sounds like we'll have, uh, another pair in the Winkler family for the pair program. That's right, little baby hatchling. We're excited for it. Um, cool. Uh, let's pass it around. Zach, how about, uh, yourself? Quick intro and your pairing. Um, let's see. Um, Well, I spent a lot of time thinking about dining out versus delivery, and I think it's interesting to see how the world's going to change over the next few years if we make the cost of delivery much more affordable. I think the way we lay out our cities changes a lot. And right now I think a lot of businesses operate in this hybrid of, you know, it's every restaurant is kind of a mini fulfillment center and factory, and at the same time it's also trying to create an experience for you. Managing those together is really difficult, but they're both very valuable. And so I think you're going to get these things diverge more and more in the future of how people think about hospitality and how they think about the brands, whether it's really for dine-in and for people joining them and having an experience versus maximum efficiency for delivery. So we spend a lot of our time thinking about how to position the world to get really good on the delivery side so that we can have more great experiences to go dine in rather than sitting in the middle of a delivery factory. Yeah, man, it's a great pairing because you don't think about it when you're going out to eat in a restaurant that there's a whole operation happening on the backend side of the business that's handling all carryout delivery orders and stuff like that. As a restaurant veteran who's worked at Chili's for 6 years through high school, The carryout line was always popping off. Obviously this was back when nobody, no robots were coming to pick up delivery orders, but it's fascinating to see how with more technology implemented, how much that side of the business scales and having to juggle those two pieces of the business. It's a really great point that you brought that up. And those will be some things I'm excited to tap into. And then yeah, quick intro, I guess, your role over at Coca. Yeah, so I started Cocoa Robotics about 6 years ago with my co-founder Brad. We were at UCLA doing robotics research, so we've been trying to make these robots do useful autonomous things for about a decade now. And the world's changing very quickly. Our kind of thesis back in the day was, we were very involved in the technical side trying to do AI research on different navigation algorithms for robots. And we kind of came away with this conclusion, like, the technology is going to get a lot better, but to make this useful and integrated into the world, there are so many things we have to solve, and we have to find a way of collecting a ton of data. And I think a lot of robotics companies were very hobbyist, super expensive devices that didn't do anything useful, but one day the technology is going to work and people are going to use them. We took the opposite approach. I mean, the, you know, the first day of business, you know, we're sitting in a restaurant figuring out how to make this robot like actually useful to a restaurant worker that does not care about your fancy robot technology. Yeah. You know, they have a tough job and they want their lives to be easier. Yeah. So we, yeah, we said, okay, well I think starting with delivery, last mile delivery in our cities is going to create a really valuable service. I think we can build something that's useful and can be useful to like, you know, retail businesses on day one. And we've been scaling it up from there. Awesome, dude. Yeah, it's an awesome problem set and it's obviously something that we've got a lot to unpack and dive deeper into. Real quick question, I guess, how many co-founders do you have? Just one, Brad. Just you and Brad? Nice, nice, man. Well, excited to hear more about that journey and the problems that you're solving. And we'll do a quick intro and pairing for you, Paige, and then we'll jump into it. Awesome, man. So quick intro. I, I'm a Marine. I went and built a private military in the first Iraq War, had a blast. Every VC in the world turned me down, so I bootstrapped my company. So I have a lot of compassion for founders and what it takes to win. I literally built my startup driving into a war zone. That worked out really well. These days I run a large early-stage fund, Outlander. I've got 6 partners and dozens of amazing founders, including Zach here. And we, we are the folks that come in and write a first check when there's zero proof anything's working. We come in and we invest in people with monumental ideas and all the characteristics to go win. So that's, that's sort of me. So I focus my time on working with amazing folks like Zach and, and many others around the country. My pairing is going to go back to experiences and, and probably ties into your kids here. So my pairing are kids and Dave and Buster's. And I grew up with a dad that took me to the arcade and there is nothing cooler than taking, I have a 4-year-old son, than taking that dude. We did it just 2 days ago and taking him to an arcade with the lamest games, the new, it doesn't matter what the games are. Yeah. We're sitting and he rushes in and we go in there and we're playing Ghostbusters and we're in and he doesn't care that he sucks right now at the Ghostbusters. But then we moved to Halo and this dude is full-on reloading. He's annihilating, like his, what his mom will not approve of this, but he's annihilating. He's going full auto on these, these aliens are dropping in. And then he wants to go to the House of the Dead. I'm like, we're not going to play House of the Dead. We're going to skip that one. You're 4. And then we go to full T-Rex mode and we're at Jurassic Park and we're annihilating this triceratops. And like, it is so much fun taking a 4-year-old to Dave Buster's or any arcade for that matter. Yeah, that's the beauty of Dave and Buster's. You can be any age, kids, adults, you're all going to have a good time there, man. That's fantastic. Kids and video games are epic. That is a good one, Paige. And Paige, honestly, man, like Sean and I were doing some due diligence on your background and we could talk for hours about your journey. We might have to do a follow-up episode just highlighting Paige Craig's journey because it's a super fascinating one. And I think this is what's gonna make this such a great episode is that perspective that you've learned from not just navigating your own businesses, but the amount of founders that you've invested in over the years, that countless number of founders is gonna bring an awesome light into, you know, really what, what, what was ticking for you with Zach and Coco specifically and what you look for in like that founder market fit. So let's jump into it cause I wanna make the most of our time. I want to start with you, Paige. So you see a lot of robotics and physical AI companies. From your perspective, what's the difference between something that demos well and something that actually works in the real world? Yeah, you know, there, there's a big and, and there's a very important difference here. There's a lot of founders out there who, I wouldn't say they're fake, but like they just, they spend all their time on the storytelling and the big visioneering. And, and, and, and you need that. Zach's really good at telling the story about why we need to go there. But what most impressed me by Zach and Brad is they were doers, right? When we met and I, I went like, I, do you remember that car garage? Like they have this like little garage in East Santa Monica and they're in there with like mechanics and they're sawing metal and they're like hacking together little robots. And you realize these guys are turning the bolts on robots, like they're making it work in the real world. And you gotta get gritty, you gotta get dirty, you gotta be on the streets. There's a lot of robotics companies that choose to build in fancy land. They build in a nice controlled, like, uh, college park, or they build in a nice factory. And Zach was like, I'm going full real world. Like, he battle tested his robots with fraternities at UCLA. He was driving them robots, show me videos of frack dudes just like beating the crap outta these robots, hijacking them. And this is the thing, like, your robot has to work in the real world. It doesn't matter if you're building autonomy for water, for, for moving, you know, hamburgers between places. You gotta do it in the real world. And so I really look for people who are deep thinkers, but paired, tied by pairing, but paired with action-oriented. Think big and execute small is one of the, those things I look for in founders. And Zach and team had that. And a lot of companies out there don't have that, but that's one of the recipes. Think big, execute small, get dirty with your robots in the real world. Yeah, that's awesome. I'd love to hear the fraternity kind of industrial strength testing there because— We need those videos. Yeah, we need the videos. I forgot about that, Paige. If we can release the videos, it seems like that would be quite entertaining. But yeah, I want to pass the question over to you, Zach. Let's do the same for Coco. Why don't you first maybe give the listeners just a quick overview of what you're building and then more importantly, the problem that you're solving. And while you're doing that, I'm going to try to tee up also a little video in the background that shows a little bit of what's going on with these robots. Yeah. So we've built the lowest cost last-mile logistics network ever, and we're operating live across cities in the US and Europe. And it's live today, delivering everything from food to groceries to packages. Medicine, right? We move anything and everything around a city, and we're able to do this at a much, much lower cost than human drivers. There are tens of millions of last-mile delivery drivers in the world, and we don't have anywhere near enough of them. And so there's this amazing increase in commerce and increasing convenience you get if you make the cost of delivery 10 times cheaper than it is today. People would stop ordering groceries for multiple weeks at a time. They could order every single day. People would get same-day delivery for a lot of their goods. And that fundamentally is going to change the way that the city actually is shaped and the layout of the city. Going back to kind of the experiences versus delivery. So, but to, you know, that's going to be a multi-hundred billion dollar business that's going to be powering, you know, tens of trillions of dollars of global GDP. And, you know, we really want to be the company that powers all of that. We zoomed in and started very specifically of, I want to deliver food for DoorDash and Uber in California because it was the fastest path to making a very, an actually useful, productive robot and getting it into the world. And there's so much to figure out to actually execute this, right? You need to get the regulatory side figured out. You need to figure out how to operationalize the fleet. How do you do fleet maintenance? How do you design the hardware? How do you actually integrate it? With a super busy restaurant that looks different from, you know, we work with Taco Bell, but we also work with super busy mom and pops that the owners are cooking, managing tables, and then managing the delivery drivers. How do I make a great frictionless product for them that makes their life easier, creates a great experience for their guests and for their customers using a totally different brand new form of technology? So the only way to do that is to find the fastest path to getting this into the world and figuring out what you need to build. And that's created a culture at this company where we had this thing at the beginning, we refused to hire any robotics engineers because robotics was so hobbyist and they just want to overengineer everything. And we were so operational and so literal, like, let's build the simplest thing possible. Let's have it break. Let's have all these things fail, and then we can figure out exactly what we need to build to have the most useful product. And that's very core in our DNA. Some of our first people at the company are all restaurant general managers. Oh, cool. You know, the guy that actually runs all of city operations for us globally now was one of our first customers. He was a general manager at a restaurant and he just quit one day and he's like, I'm working, I'm working with you guys now. I was like, we're not, we're not hiring. And he's like, well, I don't have a job. And I was like, we need to get this sale done first. And he's like, don't worry about that. We still have never sold that restaurant to this date because he left. But, you know, he knows what it takes to transition to that robot world. I don't care that he didn't have any robot experience. That guy knows how to manage your inputs every single day on a 24/7 business in kind of the real-world chaos in a service industry. So he's a prime example of a cultural pillar at this company where I want that guy to hire who's running the local market in Chicago. I want him to pick who's going to run Miami for us because he knows what it takes to be successful in this sort of business. And I think that's a very different mentality than most most robot companies would have, I would imagine. Yeah, it's a good point, like that, uh, like that operator mindset, right? Yeah, being in the, in the true seat of, uh, of the user, um, and applying it in that sense, uh, from the robotic, uh, perspective. I wanted to kind of quickly talk, touch on something that I think makes this story really fascinating, uh, with kind of like, you know, picking this very specific lane. And one of those things that, uh, Paige, you talk a lot about, we kind of kicked around a little bit in our discovery call is like this operational moat and defensibility. Can you talk to us a little bit more about what this means in practice, specifically in robotics, this operational moat? It depends on the specific business you're talking about, but let's take Koko. So if you build a delivery network and there was a time a few years ago when dozens of these existed, there's, there's a couple ways to take this. You can take the safe route of let my robot work in an easy environment. Let's just do it in an enterprise. Let's just do it within colleges. Let's do it in a limited environment. Or you can go out there and do it in the real world. And, and that's the part where like Zach was like, we are gonna make these things work in the real world, in the bike lanes, on the sidewalks, in the streets. We're going to take on this monumental task. And you look at how they have to deliver a robot, pick it up, fix it. I mean, these guys work next to the engineers who actually build and fix the robots. And for the first years, it was a noisy environment. And Zach can tell you, it's hard, but you literally have your really smart, brilliant engineers who are coding up all this AI working next to the guy who's turning the wrench on the broken wheel and the broken antenna. Next to the 3D printers and understanding how all the parts come together is really hard for most founders to understand. You are building a business in the physical world and you can't just code your way out of every problem. You have to understand the physical environment. You have to understand how you create people and process and how you collect all of this data. And like when we're dissecting problems, it isn't just, well, let's recode something. It's like, How do we have to change how the team's organized? How do we change our process? How much of this change comes from what are we going to change in the next version of the robot, or what are we going to code differently to work around that problem, or how are we going to use humans to work around that problem? So it is probably the hardest thing in the world is to build a business that spans AI, robots, right? And humans who have to make those things all work. And in the early days, you're using humans to be the glue between stuff that doesn't work, right? Like, you can't build every piece of hardware you want, so you're gonna build humans who are literally gonna be the glue that make things work before the hardware or software can, can get there. Um, and, and the other thing is, and this is tough for many founders, hardware, especially robotics, you have to have the patience to be great. There are the quick wins of you can choose the easy path, and use someone else's autonomy stack and you can take the safe approach. But like Zach went in and took the hard approach, right? And you have to have the patience to be great. You have to be willing to go out there, operate in the real world, take all that training data necessary to build a moat. And yeah, so those are some of the lessons I've seen from, from the great companies doing this. We have companies doing this in laser weapons, in unmanned ground, like weapons. Autonomous navies and spacecraft, but Zach is the epitome of what it takes to win. Cool. Yeah. Pulling on that, on that term that you just used there, Paige, autonomy stack. This is actually a question for Zach, but I wanted to read first from a blog post from Cocoa on their autonomy stack, because I think it's worth spelling out here. But you guys write, our focus is on turning massive diverse datasets into end-to-end navigation policies and simulation systems that learn directly from data. That's a, that's a pretty bold statement, right? I think that, um, when people first hear about Coco and the problems that you're solving today, they're not thinking about what sounds almost like a digital twin or a world model, um, that you're using to build what you call policies and large-scale, high-fidelity simulation systems. Um, can you talk to me about how you— I kind of want to, going back to what Paige said earlier, Zach, can you talk to me about how you combine that larger, longer-term vision for this really deep autonomy stack with the day-to-day requirements of making sure your deliveries get there and making sure your robots are operating and performing? Yeah. And it's interesting because the robots are these pink things called Koko. Because at the end of the day, we're a consumer company that has to operate in the real world and we have to the community needs to want us there. The city needs to like us. We need permission from the world to actually exist. People need to find value in these things, and people do need to want to kind of coexist with them in the cities, and they need to understand the value. So a lot of our branding, a lot of our messaging has deliberately been like— Koko was literally named off a list of cute dog names. It was the number one list on— number one name on a list of cute dog names is Coco. It was like a 2-minute naming exercise. We're like, done, call me Coco. And so a lot of the branding's around that. But behind the scenes, you need to be thinking about, okay, that's really important today to make sure we can do our go-to-market well, we can get adoption, we can get to a lot of the regulatory creation and the frameworks for this. But then what is our accumulating advantage over time?, and we're in the logistics space. Well, what matters in logistics is speed and cost. Like, those are two very important variables. How fast and how reliably can something get to you? And can you do that at a price point people can afford? And does that drive more commerce and more convenience for people? And so underpinning that is a lot of— it's really, it's three things. You have to be operationally excellent, which is kind of going back to hiring people from restaurants, hiring people from hospitality. Hiring people who know how to get their hands dirty and just work really hard every day. This isn't a demo. If a robot's down for a few hours at 7:00 PM, that is brutal for your economics. You need that robot available. Our partners are not happy if the robots are down at 7:00 PM. So you need a team across operations that just never has a bad day and can have a super reliable system. Then you need really reliable hardware, and so you need to build, and there's just nothing that exists that you can buy. So we've had to go really deep on the hardware stack to make super reliable power electronics plus all the connectivity. We connect to multiple cell carriers simultaneously. We have to be able to operate in the snow in Chicago and Finland. We have to be able to operate in heavy rain in Miami where the robots are literally underwater for large periods of time. And so it's not just like you got to waterproof it, but you also have to heat the cameras. You have to be able to blast snow and rain off the cameras. You need to be able to be submerged in salty water for long periods of time. These are unusual requirements that I did not think about 5 years ago. You think, yeah, it's got to handle some LA rain whenever we get that. But to be reliable for our partners in kind of the conditions where humans fail, Especially, right, winter storm in Chicago. Well, everyone wants delivery, but nobody wants to get on their bike and shuttle food around the city. So being reliable in those times is really important. Again, there's the operations part, but then there's like extremely good hardware, which has taken us years to get to this point where we have this kind of reliability. And then the third part is how you train these AI systems to navigate in the city, right? We, we've had a lot of breakthroughs in robotics and autonomy on the roads, you have structured road environments, you're traveling at higher speeds. This is like Waymo or Tesla that's had these huge breakthroughs recently. And then you have some autonomy on a factory floor, manufacturing facility, sometimes these campus environments or like a room service type robot, but having a high degree of autonomy in the most chaotic, dynamic, social parts of the world, which is like our city streets, city sidewalks. That data really doesn't exist. And that requires, you know, to, to, to run a high-speed and reliable service there, that requires a tremendous amount of data to understand the rhythm of the city and how to actually operate and navigate within that city. And that's just a problem that, that no one solved before, and there wasn't a dataset. And so our fleet has been able to create a lot of this data to both train the models on how to navigate and how to interact in those environments, use any of our human operations team to do all the fine-tuning and, and kind of guide the behaviors we want from the fleet, which can be different in different markets, right? Miami, no one stops for stop signs. Everyone guns it through the intersections. In Helsinki, people are incredibly yielding. You know, the traffic fine system in Finland scales proportional to your income, so people are very cautious drivers. So, you know, you want different behaviors in those markets and you need to learn all those nuances to be able to run a great service. Quick shout out to our sponsor, Defense Unicorns. This one's for the problem solvers out there. They're hosting Warhacker, a first-of-its-kind hackathon built for the defense community. No buzzwords, no slide decks, just hands on keyboard solving real mission problems with real code. You'll be side by side with developers, engineers, and innovators from across government, industry, nonprofits, and academia, all hacking for the warfighter. It's happening June 16th through the 19th in San Diego. Got a real-world problem to solve or want to join a team that does? Learn more at defenseunicorns.com/hatchit. I wanted to— you kind of touched on some of these, these things, uh, with that, that last kind of, you know, few points that you were answering with Sean. But, you know, we hear about this all the time, you know, a lot tied to Amazon, of course. Like, what with, with this last mile delivery, like, what is the most challenging piece of this? What's fundamentally broken? What are these biggest pitfalls of last mile delivery? Because I don't think people really get granular about it, but I'd love to hear like your, you know, get real specific on it, Zach, for a minute. There's a few things happening simultaneously and the reason we started with like on-demand, right? As we kind of expand into the broader last mile category, on-demand food and grocery is where we started because it's perishable. And so you're doing point-to-point trips. You have to pick up an arbitrary point A and go to an arbitrary point B in the city. A lot of the orders happen around the same time, around meal times. And you have, with hot food, you have 20 minutes, and with grocery, you have an hour to deliver it. And so point to point and highly perishable makes it really hard to run a high quality of service that customers love, and it makes it really hard to do that at a price point they can afford. So we wanted to start there. And fundamentally our business is a fixed cost business. So the idea was like, if I can get really good utilization at a relatively higher price point doing food and groceries, that will allow me to then expand that fleet to more and more use cases, which further improves utilization and further reduces our price and lets us now compete on package delivery that might be 10 times cheaper than hamburger delivery today with humans. So within the food side, you have the fundamental nature of it being point-to-point and perishable makes it expensive and hard, but it's been getting a lot worse. So in California and New York, we have legislation that has kind of added benefits and some overhead to how you pay drivers, some driver pay minimums. This not only increases the cost of human delivery drivers, it also reduces the flexibility for the platforms. And a lot of the platform's efficiency comes from the fact that people want to work for 3 hours a week. So if you start mandating minimums, Spain, Netherlands, the Netherlands, Germany, a lot of countries in Europe are now mandating that they be full-time employees of the company rather than gig economy drivers, right? So that makes it go from 3 hours a week to 40 hours a week or 20 or 30 hours a week, even if it's part-time. It just takes a lot of flexibility outta the system. So this makes the— this both restricts supply. So you have not enough drivers. We already are supply constrained in delivery. That is the number one bottleneck to growth is there are not enough drivers. But then you add this legislation on top of it, you are now restricting supply even more and you're increasing the cost of the supply. And so this is reaching a breaking point. Delivery's already really expensive. If you want to order food delivery, right, you're already paying like a 60% premium versus walking to the restaurant. And so if we can make it way more affordable, increase the global driver supply dramatically, people are going to order a lot more. People are going to be able to get things delivered a lot more. They're going to be able to get more types of things delivered, and the, the amount of use case for on-demand delivery is going to go way up. And then over time, as we get a lot of efficiencies to that fleet, we can do everything from your mail packages, returns, B2B logistics, And that's all generating more revenue and more productivity through the same fixed cost fleet. And that's how this just becomes an extremely low-cost business to operate. Yeah. Real quick follow-up on that too, just to make sure our listeners understand how your partnerships are structured. So are you partnering with platforms like Uber Eats and DoorDash, or becoming a new layer of infrastructure within them? Yeah, so we have a— we— every robot that we make has an API where people can request it to pick up and drop off things. We have a wide range of businesses using that, from grocery, retail, like local restaurants. But then we have, you know, large, great partnerships with both Uber Eats and DoorDash, and then some of the apps in Europe as well, where they could tap into our fleet as drivers as well. Okay, cool. That's helpful. Page, I got a question for you now that Zach's previous comment kind of made me think about from a sort of a broader scale perspective. What kind of unlocks do you see occurring in similar autonomy sectors or industries? Not just this last mile for customer delivery, but you had mentioned before some of your defense work and other industries like that. But if Zach, you know, kind of nails this autonomy stack, with the computer vision outputs, the data product outputs. What excites you looking further abreast to some of those other industries you're involved in? Yeah, but the good question, the cool thing is like the if has gone away. I'll let Zach do his own announcement when he's ready, but they have solved the autonomy problem, which has been a monumental 5-year journey to build technology that will It's gonna blow people away. Um, Zach was also the first company we did that had this model of, uh, what I really liked was using humans to train the AI. So there were other approaches of like, let's go raise a billion dollars and build the genius secret supercomputer highway vehicle that will do everything, or the aircraft that will do it all. But Zach's approach was like, we got humans right now. We can give 'em an Xbox controller. They can drive a robot and that driver has a copilot, which is the AI, and that AI is watching the human and learning. So he was the first company that made me open my eyes, be like, oh wow, we could do this for anything. This could be ships. So we funded a company a couple years later called Havoc, which has become the largest autonomous naval business in the world, right? And, and it's also now expanded to air and, and it's also recently expanded into construction vehicles. We have fully autonomous dozers, front loaders, like construction equipment doing this. And the unlock is that a few things. One, there's a ton of jobs out there that really suck for humans. And it's not just about like teleoperation, you know, just putting a guy in Nevada to fly a drone and drop a bomb in Iran, or putting a dude in a seat and let him run a bucket loader in Alaska, that's not good enough. What's good enough is building the autonomy so that the, the robots can do the, the, the deliveries, right? Or the, the robots can move the goods across the ocean, or the robots can do the digging and let humans do the next order of things. I think autonomy is going to let humanity do even deeper exploration of arts and sciences because the mundane work is going away. Hundreds of years ago, our forefathers had to go out there and cut down trees and mill wood, right? You didn't sit down and design a beautiful house. You had to sit there and literally mill wood for years to build your home. So, so, you know, so Zach really opened my eyes to the potential for autonomy to transform mankind by taking away relatively crappy work and letting robots do that. It also leads to these efficiencies. Like the unlock with that, with, with Koko, is if you think about the world as a network, Zach is delivering a cheaper, faster, better way to move things from A to B. And what does that mean? Well, that means that the medicine that gets to you in Kenya might be 80% cheaper, or all the people that run small businesses delivering food. I mean, what happens to the retail market or the food market when deliveries cost under a dollar. And you know, 'cause even I sit down sometimes to be like, do I want to pay $15 to have a $15 sandwich delivered to me in New York City late? All right. But like what happens when the cost is minimal, when anyone just is waking up? Like think of what happens to commerce for small businesses when delivery costs are so small. So it's the, this massive leverage that Koko is going to give on the world. It really blew me away. Yeah, and shout out to Havoc. They're an alumni from the podcast. We had Paul on. He's a fantastic entrepreneur and just really cool stuff they're doing to support some big missions. Incredible. So Zach, something that you were referencing in a previous interview, and I was doing some research I thought was interesting, was talking about going from, you know, 98% to 99% reliability is the difference between losing money and making money. Can you unpack that? 'Cause it seems so fractal, but I'm really intrigued with that percent. Yeah. I mean, this is classic in logistics. You're typically moving goods that are a lot more expensive than what you're making to deliver them. Right. And so you're underwriting the insurance policy for whatever you're carrying. And so with food in particular, you have a very small window to be successful, right? If you have if you have a system outage at 7:00 PM, that is a lot of— that is going to erase the profits of that fleet for a long time because that's— you got $30, $40 of goods in each of these orders. And so reliability is really important from an economics perspective, but also a huge part of why we're doing this is not just to drive the cost down and make it more affordable, but delivery should be a lot better. You can improve the consistency and reliability of this a lot. And I think that's a huge issue with these perishable items where you have to get it delivered in a certain window or you can't consume the food. And I think what Paige was talking about is felt across all income levels. If you look at the delivery apps at checkout and you just have this value mismatch of what it costs to how you know it's going to be, like the experience you're about to have, it just makes you feel bad, right? And so you're like, man, I'm going to pay twice what it would cost for me to walk down the street or get in my car, and it's going to be cold and soggy, and it's probably going to be missing an item like that. That is— it just has this value mismatch for people. And so it's just a hard problem, right? You have a driver that's got to go get on a bike and pick up 3 people's orders and then try to make the economics work by doing 3 drop-offs at once. You get introduced a lot of human error to the system. And so robots can be a huge help, but not just making it a lot more affordable, but just making it dramatically better for the customer. Lower error rates, less batching, more consistent on-time rates. It's one of the promises of a robotic system is we can eliminate a lot of the human error. And this is particularly intense in extreme weather, as I was talking about earlier, right? Page, we know in New York, when it's cold in New York, I mean, it takes 3 hours to get your food. It's just like, it's just this, you're just miserable. I mean, it just doesn't work. There's not enough drivers who are willing to do that job because it's a miserable job, especially in those sort of conditions. And so a robotic fleet, a robotic system operating in Manhattan would be a game changer during these kind of really low temperatures or winter storms from a reliability perspective. Like we'll keep the economy moving when nobody wants to get on the road and be moving stuff around. Yeah, it's really interesting when you think about it in that sense of like, it's like reliability is the product that you guys are delivering in a lot of ways. Paige, I want to, you know, start to close things down and make sure that we hit on something around founder market fit. You know, obviously something that you're very big on as an investor. Talk to us a little bit about Zach for a minute. You know, you talk about founder market fit and execution. What was it that you did see early in Zach that really inspired you to back him? Yeah, it's a lot to unpack there. We look for 38 things in founders. So we have 38 things we look for. Not that people have all of it, but we rate people across all 38. But a couple things that really stood out for me is that Zach and Brad— Brad's not here, but have just the most amazing relationship. And these guys met in school. Did the robot lab together, and I realized they're like this magic duo. They're both technical geniuses. Brad is extremely good at like focusing in and working with a small team to solve the most monumental problems. And Zach fully sees and understands what needs to be built, but he has the ability to sit at that like 100,000-foot level and see the big system and build the people and process to support what needs to be built on the technology. And they're both great at what they do from the tech perspective, but they were just so complimentary. And every time I pushed and asked how things were, like, they were like somehow seamlessly like working together. And that was probably that product of having just been like instant buddies in college and building things in school together. And like, they're that perfect union of co-founder that you would want in life. Like, I couldn't imagine a better duo working together. So that really struck me. And a lot of founder failure comes down to this really just dumb but really important thing, which is how good of a partnership do you have, right? That's really, really important. And then like we've been through hardship too, but before the hardship stuff, I'll just talk about the things I saw. So they had an amazing partnership that was going to be unbeatable. Like, and I, I know I'm someone who meets literally tens of thousands of founders now, 16 years into doing this. I have funded 150 comp— over 150 companies at the time I met Zak, and they had that. Second, it was very practical, like brilliant people that really could think around the sciences and hardware and software, but just practical. And this is what you miss with a lot of founders. They think about schoolhouses in theory, and they just don't put theirself out. In the real world. And so I look for that person who is action-oriented versus thinking-oriented, right? Like you need to think a little bit, but there's that old proverb of, you know, like if you, if you think too much, you're just a daydreamer, right? Right. You just, you gotta have these very practical people. And then the brainstorming, like I'm, Zach and I would have just, I met him through another founder, James Jurelecki, who I've backed 4 of his companies now, but James knew what I was looking for. James introduced us. And ever since I met Zak and Brad, I have to hold myself back because I could just spend hours talking to them. And they are— Zak is the definition of obsession. Like we have just, you know, I don't really reveal the characteristics we look for, but one of them is obsession. And Zak, even with a girlfriend, an amazing lady, his whole life, I will say it revolves around this business. And he does take the time for his family, but There is no, and like James has even told me, James is this amazing dude who is always building things, but Zak is just the most obsessive about his company, whether it's 6 in the morning, midnight, weekend, probably, I don't know, Christmas, New Year's, but he's probably dreaming about his company and solving problems. And you need a founder who has that like good relationship at home, has a good partnership with their wife or girlfriend or safely another, but who is just crazy obsessed with like a life mission thing that they're going to solve. And Zach, Zach has that. Yeah, I just, I saw with them, like he had that like ultimate obsession with making this thing work. Yeah, that's awesome. It's a theme we see it. A lot of the guests that we have on the show is this obsession, you know, folks that are, you know, working their first shift, which is, you know, call it, you know, 8 to 6 going to the gym and then working their second shift, uh, you know, for another 6 hours and then catching a few hours of sleep. But, um, very exciting to hear that. And, uh, yeah, we'll have to have you back on to try to just pull a couple more traits out of you. I know you keep it too close to your chest, but, uh, I liked you revealing just one or two there. Uh, all right. Well, we're coming up on the hour, so I want to, uh, just kind of close with one, you know, kind of outlook question for you, Zach, you know, just a forward-looking note. Talk to us a little bit about what needs to go right for this business to scale successfully over the next 5 to 10 years. Yeah. And I'll also say, when I met Paige, Paige is like, you know, because a lot of investors aren't very operationally savvy, right? They invest in technology companies. They might understand the the technology or some of these pieces, like Paige is one of the rare people that actually understands all of the different aspects that we needed to be successful. Paige was my first call when we're, hey, how do we solve the regulatory issues? I met our head of government relations through Paige. He's always been the first call on really complicated issues in our— all of our issues span hardware, software, people, operations, government relations. Right. It's always a combination of these things.. And given Paige's background, he's like the most operational, hands-on value-add investor. So we're super lucky to be working with them, and Paige has been super helpful at getting Kova where it is today. Awesome. Going forward, I mean, what we've been really focused on the last few years is getting all these things dialed so that we can actually offer a really low cost and make money doing it. That's been a combination of getting the hardware to be incredibly reliable, getting integrated with all these massive platforms and big enterprise customers, getting the fleet reliability and the dispatch software good enough, and then solving all the fully autonomous driving. We've been doing that. It was mostly across LA. We've now doing that across 7 cities in the US and Europe, and now we're going into mass expansion mode. So we're ramping up manufacturing, we're bringing Cocoa to more and more cities and really focusing on how do we build out this infrastructure in the most important cities in the world. So this is across the US, Europe, Asia, Australia. So we have a rapid expansion roadmap that's underway right now to just start integrating Koko into kind of the most important local economies on the planet. And that's starting with on-demand movement of goods in the US. That's a lot of food grocery in Europe and Asia. That's going to be a lot of packages immediately. We have huge supply hugely supply constrained in lots of parts of Asia and Europe. There are not enough drivers. So we're already starting to go to these markets and do deals where we'll deliver anything and everything and become that reliable infrastructure for the city. And I think in the next few years, I think we'll be in the most, like all of the biggest cities in the world and keep building up from there. Very cool. Yeah. Excited to keep tracking the story. Man, I wish we had more time. We got so many more questions, but We're gonna, we're gonna put a, a bow on the main discussion and we'll, we'll, we'll close with our final segment here. It's called the 5 Second Scramble. So we're gonna just pitch some, some questions your all's way. Give us your, your best answer, you know, with the under 5 seconds. Sean, why don't you start with Paige and then I'll close with, with Zach. Happy to. All right, Paige, you ready? Yep. Shoot. All right. Number 1, what was your most unexpected I really enjoyed that moment in your career. My first trip into Iraq, a year later, going to the Sheraton Hotel and having a cold beer and taking a dump when no one was trying to kill me. All right. I have to pause on that. That's one of the most Marine answers I've ever got. It sounds dumb, but it is one of the— I remember sitting in my room being like, I will remember this day. The rest of my life. I'm in a clean bed with a cold beer in a robe on a clean toilet and no one's trying to kill me for 48 hours. I was like, it was amazing. That's, that's perfect. I, I'm not sure if I have any more questions after that. Uh, that's going to be, that's actually going to be the teaser opener for the episode. That's the teaser opener for the episode. That was fantastic. Well, all right. So second question then. Um, what's a lesson from the Marines that you still rely on every day? It's all about people. It doesn't matter what technology you got. It's all about people. Great. What's a really good decision that you've made under serious time pressure? Marrying my wife. Oh, that's a— All right. Interesting. What's a book, a movie, a magazine, something, some piece of media that you really enjoy? But would never actually recommend to somebody else? Oh my God, I love Red Dawn, but it's such a bag of shit movie. I just love it. Excellent. What's something random that you're surprisingly good at that has nothing to do with work, with investing or technology or anything? I am actually an extraordinary painter and sculptor. This is interesting. We've, we've had some interesting recent conversations with investors, specifically with some creative, uh, creative talents. I think, uh, the pair program may be establishing a new paradigm here, you know? Uh, but if you need, I, I got wildly into metal sculpture because I, I originally in life thought I was going to be an artist. Like, I really, like, I love art, and my wife and I, we travel the world and we go and buy art together, but I actually make my own art too. So my house has a lot of my paintings and stuff in it. Very cool. Yeah. The, the Page Craig saga only gets deeper the more we learn. It's a lot of lawyer uncovering here, guys. Just a, just a few more, Page. You're crushing it so far. What's a, what's your favorite everyday piece of technology that isn't a smartphone or a laptop, something like that? It is my AirPods. I live in AirPods. I carry 2 because I burn them out and I walk about 15, 20 miles a day with my AirPods on. That's true. You'll take full calls just walking around with your AirPods. I interview founders. I do meetings. I do board meetings. I would do this podcast mobile if it wouldn't mess everything up. But I literally, anyone who knows me has received money from me has probably done long walks with me for hours. It's true. I got head nods from Zach. It's true. Very long, very long walks. Final question for you, Paige, um, our, our traditional question here on the Para Program: what is a corporate philanthropy or a charity, charitable effort that's near and dear to you that you want our listeners to know about? Man, there's an amazing one which I can't talk about because of the unit that we support, But US Vets is also an extraordinary group that spends almost all this money housing veterans across America. I was lucky enough to serve on the board for a long time, and they're based out in LA. Incredible group. Awesome. Paige, thanks so much. Over to Tim and Zach. Yeah, good stuff, Zach. You ready? I'm ready. All right. If Koko Robotics were a restaurant, what would be its signature dish? Oh man, um, Street Tacos. Hell yeah, well said. Nice. Last good in delivery too, don't they? That's right, it just— you can't beat Street Tacos. The, the company was created by eating Street Tacos right out front of our office. Okay, we have that every single meal. Uh, what are a few, uh, key roles that you guys are hiring for over the next 3 to 6 months? Lots of engineering, hardware engineering, software engineering. Anyone that's got really good product, product mind and operations systems mind across operations, software, hardware engineering, I'm, you know, please, please reach out. Is a lot of that out in LA or whereabouts are you guys hiring? LA. We have an office in the Bay Area and we have a team we're building out in New York. Very cool. What's one thing about the culture at Coco that you would say surprise a candidate? Uh, probably how like incredibly hands-on every single person at the company is. I mean, if there's an outage, everyone's getting in their car and they're going to go deliver food, which has happened before. Uh, the whole company has to stop and go make sure the food gets delivered on time. Sometimes people think they're going to a robotics company and they're going to be a fancy AI lab, but no, the real world is not that neat. That's so gritty. I love that answer. What's been the hardest real-world lesson that you've learned building this business? Oh man, definitely Miami. That place, I love Miami, but that place is brutal. The way people drive. I mean, there's just the way people drive, the amount of rain and the saltiness of the rain. Like we thought Chicago was going to be hard and it was, but Miami, getting Miami reliable was next level, next level difficult. What's, what's one metric that you watch most closely that tells you the business is is working? Um, the high— the most important number is probably like our high-quality delivery rate. So we— that's a ton of different factors of how we measure quality, but that needs to be above kind of human courier baseline, and all our new markets need to be operating at our baseline, and then we move that baseline up every month. Very cool. What was your very first job? You know, it was actually, I was a tutor for pre-calculus, which is, and this is, I grew up in Menlo Park in the Bay Area. So this is like the most Bay Area thing ever. But there was a startup founded by a bunch of high schoolers and we were making online tutoring videos and somehow I got assigned pre-calculus, which is the worst math subject. It is like a random collection of like all of the things you need to know in math before you go into calculus. And it is the biggest textbook. It is the most miserable course to teach. And so a whole summer I had some sort of tablet and I'm writing out the entire textbook of pre-calculus, which was terrible and traumatizing, but it was a fun— we were in this little house in Palo Alto and there was a bunch of kids trying to build an internet education site. So that was a lot of fun. That's crazy. I just had an episode yesterday with the founder of an AI CRM startup in the Bay Area. And his first job was pre-calc tutoring, and he was just basically undercutting the market of like anybody that they're doing like $14 an hour. He was like, I got you for $12. So ironic. I wonder if you know this guy. All right, so you spent years kind of thinking about autonomy. What's one everyday task that you wish you could have fully automated in your own life? Oh man, um, I'll say besides delivery, because I order a lot of delivery, and, uh, I guess I've already had that one automated, uh, which is nice. Um, my, my girlfriend's a great cook, and so she, she, she cooks, um, but of course then I have to clean. And she's Italian, so the way she cooks is by like spewing ingredients all over the place. Um, It's part of the process. You know, you can't mess with the process. And then her dad comes over. I've literally held a bowl before for him while he is cooking and he has tossed olive oil into the bowl from across the room. So, so probably, probably I'm, I'm very excited to have a humanoid that can properly clean the house after, after an Italian meal. Preach, man. Yeah. Automated dishes. That's, that's a home run answer. Uh, what's the weirdest real-world edge case that your robots have ever had to handle? Where do I start? Um, it's all edge cases, man. One of the first, one of the first things where we're like, okay, this is something we're gonna have to get really good at. Um, this is in 2020 and 2021. We were delivering, we were delivering a lot of Chick-fil-A sandwiches in Hollywood.. And there was these, all these homeless encampments in Hollywood and every morning they would move and you didn't know where it was going to be, but they'd take the whole street and it was always a different street. And this is like the most important customer. We're like, in Chick-fil-A, you got to deliver on time or you're fired. So they take quality super seriously and we would never want to ruin one of those sandwiches. And so we basically had to have a scouting robot that would go out at the beginning of every morning. Just to figure out where that homeless encampment was that morning, because the city kept moving it, and so it would pick up and go somewhere else, and it would cause us to have to reroute because we couldn't drive through it. And that reroute would add a few minutes, and that few minutes we could not tolerate. So we had to build a whole real-time system for homeless encampment tracking in the city. And that turned out to not be a corner case. That is pretty common. But yeah, there's a lot of things like that, active crime zones. We've had a robot got run over by a car from somebody doing like a smash and grab into a restaurant, right? And like went through the robot. There was another robot behind the robot that got hit, just like watching his brother get run over. And then the people ran in, you could like get all, you can watch them all just run out with a bunch of stuff. We see a lot of crazy stuff. My gosh, that's, uh, that's amazing. We, yeah, we gotta do a full episode. We've been, we've been, uh, the robots have been shot before. We have found a bullet inside the robot. Um, it was delivering by USC and we just caught a stray bullet. Driver. Yeah, better than driver. Paige, it did, it did deliveries for like multiple days and it came back in for service and we're like, all these USC kids are just like, yeah, it's fine. No one's reporting it to us. Oh my gosh. We also got the Cocos that you sent to war. We did send robots into Ukraine. I don't know where they went, but we put a Red Cross on them and sent them in and haven't heard back. Train them in LA. Train them in LA, then send them to Warsaw. Yeah, hopefully they provided some supply drops before they— before they got blown up. That's cool, man. So many like business ideas within the business. It's really interesting. All right, we'll close with the last one. A charity or philanthropy that's near and dear to you. We do. We've actually been doing— as funny as Paige brought this up, we've actually been doing— we deliver around the VA hospital here in LA, and actually we had to deliver just to the gate. And we've actually built a relationship with them of how we can do a lot of useful services within the VA. And that's actually turned out to be a really cool partnership where we couldn't go in and now it's like, well, they actually needed to get a lot of goods into the facility and they need a lot of goods to move within the facility. So we've actually come up with a really cool partnership with them of how they can use the robots. And that just happened organically from the fleet driving around. That's been a really cool project. Very cool. Yeah, we'll give both of those a shout when the episode goes live. But Zach, Paige, just wanted to thank you all for spending the time with us. This was awesome. This was probably one of the more fun episodes I think we've hosted in a long time. So thank you guys for bringing the energy and explaining a little bit more about the future of autonomy and on-demand delivery and scaling in this space. Really, really fascinating conversation. Thanks for joining us on the show. The podcast. Thanks, guys. Yeah, thank you, guys.