Ep44: Is Hyperlogistics the Future of Autonomous Delivery? w/ Canon Reeves, Co-Founder & CTO @ Pipedream Labs — Product Market Fit podcast (startups | tech | AI | growth)

00:00 Canon Reeves we kind of have this like node capture strategy. Like the more nodes we have, the more physical places we wire up for autonomous delivery, the larger the wedge we have kind of in the future autonomous delivery market. 

00:08 Mosheh Poltorak Hello and welcome to product market fit, a show all about startups, technology and growth. I'm your host Moshe Poltorak and you're in for a real treat today. I couldn't be more thrilled to share this interview I did with Cannon Reeves, CTO and co-founder of Pipe Dream Labs. His company is revolutionizing hyperlogistics and shockingly, they're actually bringing their futuristic vision to reality with announced partnerships with the likes of Wendy's and the city of Atlanta to name a few. My excitement about this episode isn't just nerding over this Jetsonesque technology, but I find it super inspirational to see founders taking on massive problems with equally massive ambitions. I want to give a shout out to my friend, Michael Shore, who's an investor in Pipe Dream for introducing me to Cannon. Heads up, this is the first in-person interview that I've done for the podcast. So it's going to be a bit different than usual. And unfortunately, there's a bit of sound issues that I'll resolve in the future. If you want to watch this interview instead, the full episode, like all our episodes is available on our YouTube channel at PMF pod. The product market fit podcast is brought to you by growth.co, that's growthwithoutto.co. Growth offers fractional CMOs paired with best-in-class digital marketing execution to support early stage startup success. With a focus on seed and series A companies, growth has helped a number of SaaS, digital health and e-commerce startups build their go-to-market function and scale up. To learn more and book a free consultation, go to growth.co, that's grwth.co. Without further ado, we have Cannon here. Let's talk about Pipe Dream.

02:02 Canon Reeves Can you tell us what you're building? Yeah, Pipe Dream is one of those fun companies to pitch because every 60 second pitch you give breeds more questions. But basically we are building the future Piper logistics. And the way we define that is an end state where you can send and receive any items in single digit minutes from your home. It's just sort of this idea of like ultimate convenience. To achieve that and build that, we are building underground delivery robots. And so basically we have taken standard utility pipes and retrofitted them with rails and robots that allow us to do like completely invisible high speed

02:33 Mosheh Poltorak and low cost delivery in cities of any scale. Fascinating. So robots delivering things underground in pipes, hence the name, right? So talk to me about the logistics landscape and where you guys fit in, right? We know first mile and then you have the entire delivery system and then you have last mile.

02:49 Canon Reeves Where does Pipe Dream fit in? Pretty firmly in the last mile. We don't compete well with like semi trucks because semi trucks are super efficient per package. The cost of that truck and that delivery is pretty low because they can fill so many of them in there. Where delivery really breaks down as the last mile because you have one person carrying one package or one burrito in kind of like a door dash type example. So that's really where we shine,

03:08 Mosheh Poltorak is things that are like one driver to a few packages. Okay, so the end state would be delivering things directly into people's homes, right? That's the vision. Yeah, that's the dream.

03:17 Canon Reeves And that's how far out? You know, closer than people think, it's really easy to build this system in when you're doing a new build community or a new apartment complex. And so what I suspect is in the next three years, we'll be doing actual delivery into homes. It will take a long time to scale that. Obviously you're not gonna go retrofit into a single family home. That looks a little bit different, but there definitely will be like new builds that have this system because it's really like an incremental cost to the developer to add this in. It doesn't increase the cost per home very substantially under 10K per home. And they're already laying all these other utilities. And so basically they can, while they're like trenching for these other things, just make the trench a little bit bigger and include our pipes. And a lot of the costs just come from the portal hardware that'll come down over time.

03:57 Mosheh Poltorak And it's a great new and very unique amenity. Got it. You mentioned the cost of the developer. This is, I'm assuming high capex type investment whenever you're laying pipes on the ground, you're building robots. In this model, when you're delivering to consumers, who's paying for the service?

04:13 Canon Reeves What's your business model? It looks a few different ways depending on the install type. Because at the end of the day, this is utility. I don't know in the end state if we wanna be the ones owning the pipes necessarily. We definitely wanna tie sort of our business to the incremental deliveries that the robots do. Sort of like just making a couple of pennies here or there every time a robot makes the delivery. In terms of pipe ownership, it can go a number of different ways, whether the developer owns that or if it's utility, company wants to come in and lay all that pipe, they could own that and do like a leasing model. It looks a few different ways, but at the end of the day, we make money when we do more deliveries.

04:46 Mosheh Poltorak However, that actual fine details, you know. You had a tweet thread a while back and you described the nuance of your delivery network versus utilities because utilities are one to many, right? You have the water plant or the electric generation happening elsewhere and it's being distributed all over to the end users. You guys are gonna be multi-user on the front end as well, right? Because Amazon and Wendy's and whoever wants to deliver things to many houses, so it's many to many, right?

05:14 Canon Reeves How does that change the design of this? Yeah, it makes bootstrapping the network much harder because you have way more parties. If you think about it in terms of just stakeholders to make something like this happen, the very first is the city, right? They're more of a gatekeeper than a stakeholder. We never anticipate the cities to be paying for this. They have enough other things that they need to be focused on that this is really something they just need a permit. So this is the city's allowing us to do the right away. You have the supply site users, which are your Walmarts, your Amazons, your food, DoorDash, UPS, those players. And then you have kind of the end destinations, the final nodes. And when you think about building a network in a city, the beautiful thing about our product in our system is that as more of those players get added on, the value of that network increases. So it becomes more and more enticing to plug into that, especially if you're on the demand side like a developer. Our strategy has been we have to create products that are valuable to those parties independent of the full network. So the Wendy's example, we want Wendy's to be able to send your order directly into your home. That's the end goal. But it would take forever to get there. And if you try to do just that from the beginning, I don't think it would work. What we've done is instead say, let's chop up our product and offer a version that's kind of like a bank tube for foods. We call it first inch, where you're going from the back of store in a Wendy's to the parking lot to do pickup. So they're already doing like mobile pickup for people using their app and DoorDash drivers. And we can take that and make that process more efficient, pull people out of the drive through so they can fit more people through and just build a better customer experience. It's kind of like digitally first. It's kind of like a self-checkout for fast food. But by building this product, if that can be independently useful and scalable, the scope of that network is just that parking lot. And so you can go do that everywhere. And that can be a product that makes its own money and scales on its own. And then later on, as you start to get more developers and larger city installs planned, you can connect up these existing supply-side nodes because you're already providing value to them

07:02 Mosheh Poltorak and already installed. That's a really interesting way to break out your rollout strategy so that you can deliver value along the way. And congratulations on that deal with Wendy's. That's massive. How did you land that deal before we get into the specifics of some of the other parts of what you mentioned? But I'm really curious about startup. You don't have any revenue yet. I'm assuming this is your first revenue generating deal.

07:22 Canon Reeves How did you go about doing that? Yeah, we've had people ask us, who did you know there? And I wish we knew somebody there. No, we didn't. We just cold emailed and got in touch with the innovation team. Their innovation team is amazing. And really everybody across that org, we have worked so deeply with them across every function of the product and everybody there is amazing. So we honestly got really lucky to have them as the first customer. They've really helped us iron out a lot of the details. I'm a big fan of building with your customers. If we'd gone in and said, we're gonna tell you what you want, I don't think it would have worked well for anybody. But we went into them, shared the same vision, and collaborated together

07:54 Mosheh Poltorak to make the best product possible for them. So how do you structure a deal like that? Is it, we'll figure it out as we go along. Do you have end state goals that you're working towards in order to, with milestones,

08:03 Canon Reeves how do you structure something like that? Yeah, this is definitely a pilot. All of us are really aligned that this is kind of a test. The hope and the goal is that we then go do more locations and everything we're doing is sort of with that lens of like, we know internally what we want to hit and what we want to see. At the end of the day, it's increasing revenue for them is the goal. But yeah, assuming things go well, we'll want to turn around and do more.

08:24 Mosheh Poltorak But there's a lot of things between now and then. So the pilot starts with how many locations and when are you gonna be live? It starts with one. It goes live at the end of the year. Fantastic, okay. And then hopefully that's successful, then they're gonna expand it geographically to more and more locations. Yeah, yeah, that's the dream. So I'm a customer. I come to Wendy's, instead of a drive-through, it's like a kind of pickup locker sort of thing.

08:46 Canon Reeves Is that how you envision it? Yeah, yeah, there's a few different versions or like use cases that you would have with it. But let's say like you're a customer and you order on your phone ahead, which is like, in my mind, it's like a faster than better experience. And so basically right now, if you do that, though, you have to either walk into the store and find your order or go through the drive-through, which at that point, just order on the drive-through. So there's like, there's some friction in that experience. And so with our product, basically you place your order, you come in, you pull into any of the portals, there's a number of them there, kind of like a Sonic drive-through. You pull in, you scan a QR code on the screen and your order gets sent to you. And like 30 seconds to a minute, your food is there. The door like magically opens and this kind of mechanism comes out to hand you your food. And like, we really wanna make it just this like unique and magical experience. That's just something nobody's ever seen before. And then yeah, you get your food and then you go. And you don't have to talk to people, you don't have to like wait in line and keep moving up. Like you can just sit there playing your phone

09:35 Mosheh Poltorak and get your food delivered right to you. So part of me, I've never been to Sonic. How is it different from a Sonic? Do they have like pneumatic tubes that they're using? They have people that bring their food out to you. Cool, okay. But since I brought up the pneumatic tubes, we've all seen them either at a bank or sometimes Costco will have them. Is this system built to model off of that? Like how's that different?

09:56 Canon Reeves Do you think about that system? Yeah, so kind of where it differs, the pneumatic systems don't maintain the orientation of that tube. So when it goes vertical, it goes straight vertical. When it goes horizontal, it goes straight horizontal. And if you imagine like your food order being twisted and turned like that, it doesn't really work. There's also fundamental differences in how big you can make that pipe because it's pneumatic. There's also like worrying about sealing things. So our strategy is totally different than pneumatic. Basically what we do is we have this standard rail that we lay throughout the pipe and the robot drives on that. And then there's kind of like a small elevator that comes down and grabs the order out. So the robot's always staying in the same plane and then a separate robot, as you can think of it, lifts that tote out and moves it around. See, to put it more precisely, it gives you a lot more flexibility with how you construct networks

10:38 Mosheh Poltorak and how you install this and where it goes. Right. And imagine with the way pneumatic tubes

10:43 Canon Reeves are creating those vacuums, it's pretty limited in what you can do. Yeah, yeah. Especially distances, right? And so kind of the cool thing about our product that is sometimes surprising to people is that that product that we're doing for Wendy's, which is just in a parking lot, is the same exact one we're doing at our pilot just outside of Atlanta that's like a mile long. So none of the tech changes. It's all completely modular and all interfaces with each other.

11:03 Mosheh Poltorak And it doesn't work that well when you scale that to pneumatic. Right. OK. So it's an underground tube. You've got tracks inside the tube. You've got a robot that's going on those tracks and then you've got robotic elevators. Yes, we call them portals. That'll grab the cart or the robot off of the tracks

11:20 Canon Reeves to that last access point. Yes, to be more specific, to grabs like a tote, like just a plastic tote that holds your order. And then one of the cool things you can do is like we can store orders above ground and kind of an air conditioned area. So if like that order gets done before you're there,

11:33 Mosheh Poltorak we can kind of have that cached. Very cool. So talk to me about that.

11:36 Canon Reeves You mentioned the Atlanta pilot. What's the use case there? Yeah. So this was sort of the first build out we've done. We really wanted to see, to validate a few things. One, validate the tech, but two, validate that we can work with cities and get this permitted because we had no idea how they would react to this. And it turns out they are super friendly to this. We're just installing pipes that they're very familiar with. It's not under pressure. It's not really a lot of risk associated with it. And so it was pretty easy to permit it. It took like two weeks to permit this one. But it's a three-quarter mile track going from a retail center we have inside of a strip mall all the way to an office building. And so it goes directly inside the office building. And the cool thing about the robot is it can go in pipes, but it can also go in this overhead ducting that we've created. So it goes into the basement of this building, into this ducting, and then drives throughout the building to a portal.

12:17 Mosheh Poltorak Awesome. Yeah, it's a fun one. What's the consumer response been so far? There's a novelty aspect to it.

12:23 Canon Reeves But beyond that, how are people reacting? Yeah. So that one has not gone live yet. So we are in the middle of bringing that up and getting on a plane in like a couple hours to go do that. But that pilot is less of a business validation pilot. It's more of a like technology validation. It's city relationship validation. So people want it like nobody wants to be the first one to do it. They want to see somebody do it. So that one we've really focused on. Let's let's prove that we can do it. And then let's go to the next ones and make them like super viable.

12:46 Mosheh Poltorak Yeah. Breaking it apart, you have that first leg of delivery, getting it out of the Wendy's into the parking lot. And then with the portals, it could be a customer that's driving up to the portals to grab their package. But it could be a door dash driver as well, or eventually could be a drone or whatever, picking it up there, bringing it to the house in theory, right. That last mile. On the other end, you're building out the in-home delivery network for new construction developments. And then the vision is that eventually they'll connect with those long tubes that are that are citywide.

13:17 Canon Reeves Yeah. Yeah. And there's a few different ways it looks. And so like kind of the thing I want to emphasize on this, too, is like, this is such a hard, messy, if you will, business that I don't think there's like a clear, oh, this is the one way that does it. It requires a lot of flexibility on the go to market and the tech to kind of like go the path. The water flows, you know, we're definitely pushing for different kinds of installs like citywide. But, you know, we have like kind of the insight we want to get to, and we have the tools to do it. It's just kind of a matter of like the relationships and the business development and the projects that come about.

13:48 Mosheh Poltorak They get us there. Yeah. It's super cool to be going on a path with such an ambitious vision and long term horizons. How do you do that day to day? Just from a founder perspective, keep yourself motivated because, you know, you're not able to ship product to customers that quickly compared to, let's say, a software startup or whatever. You want to see that end state. But also as it relates to investors, you guys raised venture capital. You're going to need more venture than you're going to need to raise. And how do you pitch that where the returns are going to take a while?

14:17 Canon Reeves Yeah. You know, usually we get people who are investors are all super bought into the long term vision. They understand that, you know, there's some generational companies that just take a decade or two to build. We're really ambitious, though, in the timelines, because fundamentally we've taken autonomous delivery from a tech problem and turn it into a financial problem. You know, self-driving cars and drones are very difficult. Technically, you have to work with the FAA and certifications, but like the reliability that you have because there's a safety element where you're like interacting with people in the public. We're totally different. Our robots are really dumb. They're very simple. Like they're not easy to create, but they're not hard. Like in the same way a drone is, they're hard in the sense of like we have to make them last a really long time and work really well and be really reliable, but also be cost effective to produce. So everything we do, we design for mass manufacturing. Like everything is sheet metal. Everything is made to be scaled. So we kind of view our ramp up a lot differently than like, you know, some of the deeper tech companies, because it's like we know how much the pipe costs. You know, if you're working a great partner, they know much how much volume they're going to be sending through it. So it's a pretty easy financial equation. Then you just have to get that pipe financed and do a longer term loan

15:21 Mosheh Poltorak and pay it off. So you mentioned drones. They've been promised for a while as that solution for last mile delivery. I know that there's been pilots. There are some active pilots, but it's still not widespread. Do you see this as replacing the need for drones? Will underground delivery coexist with drones for that last mile?

15:38 Canon Reeves What do you think? I think it'll coexist. You know, there's definitely things holding back that industry. Regulation being a large one of them. And they're great, great companies that are doing really good work there. And we like our internal views that drones are going to happen at some point where drones break down. Well, not where they break down, just one of the difficult problems with them is like, how do you get the worker in that Wendy's to load that drone? And how do you get them to do it safely? And like, how do you do it without them having a ton of technical knowledge? You know, when you work closely with these QSRs, they care so much about operational efficiency. Like it was a huge thing to figure out where exactly we wanted to put this portal inside the store just so that it's operationally like right where they want it to be. And so our view on drones is we have this first inch product that is essentially like wiring up a store for autonomous delivery, because either it's going to be us taking it to the end or it's going to be us handing off to another modality that's going to take it to the end. But at the end of the day, that store just has to interface with us. And if they can do that and do that scalably pre drone scaling, we're in a really good spot to be sort of like the USB port for autonomous delivery. So I think drones might be a thing. And when they do and if they do, you know, where they are ready for it, we will work closely with them. And they don't have to worry about interfacing with all these stores. They just work with us. We develop that kind of handoff protocol. And then they can do their thing and the store can do their. So once you get the product outside of the building through pipe dream, any modality has access to it at that point. Yeah. And then it's just basically changing what that portal, what the end node looks like. But we view it as like we kind of have this like node capture strategy. Like the more nodes we have, the more the more physical places we wire up for autonomous delivery, the larger the wedge we have kind of in the future autonomous delivery market.

17:09 Mosheh Poltorak Cool. We have a thing about defining acronyms here. You said QSR is that quick serve restaurant? Is that what it is? OK. Quick service restaurants. So what about the parallels between what you're doing and Elon Musk's Boring Company? I'm sure you've heard that many times before. First thing that comes to mind, we talk about digging holes on the ground and and moving things through it. How is it different technologically?

17:29 Canon Reeves And where do you think that that plays into the future of logistics and transport? Yeah. You know, Boring Company is so interesting because, you know, they are this like huge high profile company doing underground stuff, but polar opposite in terms of what they're doing. So Boring Company is innovating on the actual tunnel digging method, specifically with these really large diameter, like 12 meter tunnels. We take the opposite approach. We say we do not want to innovate on construction, at least to begin with. We want this to be a system that's viable with current methods. So every time we're doing a pilot, every time we're doing an install in a location, we know that there are at least five contractors we can call up in that area that could come and install our system. We'll have to help, you know, some pointed stuff. But like because your pipes are standard, they've just done this stuff. It's not really that big of a deal. Whereas boring is like they're creating this whole new machine. They're doing these things. We're very much a tech company, not a construction company.

18:17 Mosheh Poltorak What about with regards to like how the cars move through their tunnels? Is it just using tracks and wheels? Is it electromagnetic? How are they transporting things?

18:25 Canon Reeves Do you know? I think that their product works in a few different ways. So some like I think in the Vegas case, it's kind of like a track, but it's more like a concrete, very well defined road for the tires. But I know that they're also kind of thinking about Hyperloop long term. And I really love their thesis of like being able to essentially create new roads without taking up land on the surface, because like traffic is absolutely a problem. There is space underground for things. And, you know, kind of when I think about our product and interfacing with the city, one of the things cities like about us is that we're taking traffic off the roads. It's like you're turning this like giant two ton vehicle that's carrying a burrito into this tiny little robot that's invisible underground in the small 18 inch pipe. So it's like I really love this idea of taking traffic off the surface

19:05 Mosheh Poltorak and just putting it underground. Yeah, it's ridiculous if you think about the fact that our our roads these days are you have these two ton metal machines with a human

19:14 Canon Reeves delivering a small bag of burritos. Yeah. And it's crazy because the value of that service is super high. You know, I do all this stuff all the time. I love being able to just get my stuff delivered to me instantly. But, you know, I don't think that it's the most optimal solution. You know, I think there's room, but it's just, you know, autonomous delivery has just had so many setbacks and just there's so many hurdles to overcome. I think that this kind of new angle we're taking could could really like push ahead and make a lot of ground.

19:39 Mosheh Poltorak Yeah. I think that until we have in-home replicators where you can just print on demand that burrito, I think that innovating on the delivery method is necessary. And what you guys are doing is fascinating in that space. You mentioned Hyperloop.

19:52 Canon Reeves Can you talk a little bit more about that? Yeah, honestly, I'm so so naive and I don't know much. But my basic understanding is that you like create a vacuum in these in these tubes and like bring the air resistance down to zero and you can go super high speed, super energy efficient.

20:03 Mosheh Poltorak But I've known I don't know anything about it. That's a future state. Yeah. Got it. The pipes. You mentioned that you're using standard pipes. Can you retrofit existing pipes? Because I know that there's especially in older cities, there's a ton of underground. Just no man's land, right? Things that have been laid for decades and decades of that are not used.

20:20 Canon Reeves Can you retrofit any of that? Yeah. You know, we took a strategic step to separate the thing the robot drives on from the pipe. It's just the rail and the rail is not affixed to the pipe for a very good reason. One, you can uninstall it if you ever need to. So you ever need to pull the rail out and change it out. It's super easy to. But two, we can install that pretty much anywhere. So as long as there's a pipe that's larger than the sort of like 18 inch internal diameter, you could put this rail through it. Now, retrofitting comes with its own issues in terms of like cleanliness. I'm like, is the route actually helpful? But like from a technical standpoint, we're super well set up for it.

20:51 Mosheh Poltorak OK, cool. What about standards? You guys are creating not just a new company and new technology, creating a new space, a new industry and a new standard. Right. I was listening to Ben Horowitz talk on a podcast and he was talking about the early days of the Internet and how we look back now and, you know, it was inevitable that the Internet became what it is. But it wasn't in the early 90s. Right. It could have been a lot of different types of networks. We got lucky that we got a open network that became universal as opposed to, you know, if Microsoft had owned the network or Oracle or whoever. How do you think about standards in building this new technology so that, you know, there's going to be competitors? I think there are already a couple of folks that are claiming similar things or building similar technologies. Are you working together with them or with industry bodies

21:37 Canon Reeves or regulatory bodies to create those standards? The interesting thing about laying actual pipes is where each competitor is different is basically the tunnel size and the pipe size. The pipe size is downstream of the target market. And so we're the only company right now that I know that's like really targeted to go to the end home. And for that reason, we've sized on our pipes to 18 inches. Every other competitor is much larger. Will there be competitors that are close to our size as we become more successful? Absolutely. You know, I'm sure TVD going to have a collaboration goes and how those plug in together. You know, we've definitely taken the approach of like this has to be an open network. We're not going to sign any exclusive deals with one grocer or one food service company like it is made to be an open network that all these players can plug into. We also don't necessarily see ourselves owning all of the pipe and all the network. We just want to operate the robots. I think when you get into the world of like allowing other companies to operate robots in those pipes, you run into like the risk of it operationally failing. But I think there's other ways that you can standardize within the industry, particularly in the autonomous handoff. And so what we've really tried to push in our starting to spearhead is this effort to standardize this one container, kind of like a shipping container for individual orders that can be handed off to these autonomous modalities. So like I think that's where a lot of the standardization will come in.

22:47 Mosheh Poltorak Great example. Exactly. I was thinking of the shipping container. You know, somebody argues the biggest revolution technologically of the last hundred years and what it facilitated because standardization allowed for all of that downstream commerce to happen in a much more efficient way. Totally. Yeah. So you're certainly not competing against other modalities, and you're also open to potentially collaborating on whether it's the pipes themselves, the robots, TBD. Is that right? In terms of collaborating with competitors, collaborating in a future where there will be multiple underground delivery systems, and it would be shame if as a consumer, you can only buy from a certain number of retailers because your zip code, right? Like I'm locked into Comcast and that's all that's my only choice, right? Yeah. Because that's what's been piped to my zip code, right? As opposed to an open network where everybody is cross compatible and can work together.

23:36 Canon Reeves Yeah, we'll have to see how it plays out purely from a technical standpoint, because I think ideologically, we're super aligned to that operationally. Technically, it gets hard just because the robots need to move at such high speeds and at such high density. You don't want to introduce too much risk. But I think we're super aligned with giving consumers choice, because I think it's like our incentives are to drive down the price of delivery at the end of the day. Yeah. Easier said than done, right? Absolutely. Yeah.

23:57 Mosheh Poltorak You mentioned they're not offering exclusives to some of the supply side partners. But going back to you mentioned this as a financial problem, right? And who's going to pay for the pipes, who's going to pay for the infrastructure, who's going to put up that money, even the homebuilders, even though they're getting some of the value, I would see that most of the value gets accumulated at the top with the supply side, right? So could you do maybe not a full on exclusive, but some sort of like preferential deal with, let's say, the local Amazon distribution center or Walmart or whoever. Maybe they get a six month head start or whatever it is so that they finance it. And then other supply side partners can come onto the network.

24:31 Canon Reeves Is that a model you guys are looking at? Yeah, I think there's so many fun financial engineering things to do there that are beyond me. A simple mechanical engineer. I think where we would try to align things is more of like a leasing and ownership model. So we would want people we would want to incentivize partners to finance sections of pipe, but basically say you guys get a percentage of the revenue of all deliveries to go through that because you put up that capital. So then it's a lot easier for them to do the equation of like, OK, well, if we're putting this much volume, we know that just pays back. But also, if these other players come and put in volume, it just increases the money they're making off that pipe that becomes an investment for them. So I think the real like in the real goal would be make the pipes this financial asset that's easy to invest in similar to fiber to where all of these kinds of independent capital partners can go and invest and create these special vehicles and do whatever they need to do to create that pipe.

25:21 Mosheh Poltorak OK. What's been the biggest change in your thinking since you started? You've been at this for what, three, four years?

25:27 Canon Reeves That's it. OK, two years. So I'm on the technical side. So it changes some things, but we always approach this as just creating like the most modular and easy to install network as possible. There's definitely been a change in balance of like where we put the complexity of the network. So we've got a competitor in the UK called Magway, and they've taken this approach of turning the rail into essentially a long linear motor and having the robots kind of just be a cart with a magnet on it. They're very simple. There's not really any complexity in the robots, but all the complexities in the infrastructure. Our approach has been put zero complexity in the infrastructure, because that's the thing that the cost per foot of that infrastructure is the bounding metric for how quickly you can scale this and put the complexity in the robots. Where we ended up landing was a little bit in between both, where we have this kind of dumb, unaffixed rail in pretty smart robots, but also pretty dumb, like not very complex. But there's definitely more complexity into that network than I into the infrastructure than I expected in terms of like as a business. That's how our thinking has changed. I don't know how much has shifted. We've been pretty keen on this vision and direction and just plugging away at it. It's a pretty hard business to build out because you don't know a lot of the requirements until you start working with the customers, but you can't really start working with the customers until you're far enough along to validate that you're real. So I would say we're kind of now at a phase in the company, and this year really, where we're working with customers enough to really flesh out the finalities of the network and the design. But yeah, I'd say it hasn't changed a ton.

26:49 Mosheh Poltorak We've been kind of plugging away at it. OK, very cool. What about in your journey? You founded a couple of companies before this, a robotics company, I believe, that was acquired and another company. Tell me about your journey, how that led you to where you are today

27:03 Canon Reeves and what you've learned as an entrepreneur. Yeah, you know, I've been doing robotics since I was like 14, and I've always wanted to be just the best robotist I can be. In college, I got introduced to entrepreneurship, and I dropped out and started a robotics company. And this robotics company was making this modular education platform to teach kids how to code and how to CAD and 3D print. But it was kind of like Lego in that it was all just this super modular system. And we got picked up by a large company, the largest company in the industry, because the education industry is not fun to sell to, especially selling hardware during COVID. But we had a good deal with them, and basically I got to join this company and redesign my entire product to be mass manufactured. And so I've kind of gotten to go through like three modular system developments from scratch in that process. Creating a really good modular system is really hard, because you basically have to push your modular rules to be as simple as possible. So for Lego, those modular rules are like the brick interface. You can always create all different kinds of bricks, but that interface stays the same. So coming into pipe dream, it really set me up to view the network that way. And to be really particular in the early days, we spent a good year and a half just getting those modular rules right. Because once you get those right, you're off to the races. You can develop so quickly and create such a great network. But if your rules are wrong, it becomes this like five to 10 year horrible decision. And I think that it just taught me coming into pipe dream that the best thing you can do developing a modular system is be very, very patient in the early days while you iterate on those modular rules. And we created a ton of prototypes, very quick prototypes, getting those modular rules just right, talking to customers, figuring out what the actual needs are. Because we didn't know all the constraints when we started developing. But yeah, I mean, now we're at a point where those modular rules are really set in place. And we've got a lot of mechanical engineers

28:42 Mosheh Poltorak who are just cranking away, making actual hardware for it. You know, most of the conversations I've had on the podcast are with software founders. And, you know, in software, you're thinking about backwards compatibility, maybe a couple of versions. But when you're talking about laying pipes, you have to think about what we're doing now. We're going to live with that decision for 10 years, 20 years. Yeah. You got to make sure future you doesn't hate current you too much. Yeah. You guys have a fascinating rule as it comes to hiring. I heard Garrett talk about the idea that every new person has to be better than the last person that you hired. And you've managed to attract some folks like Justin, co-founder at Drizly. He doesn't need to work. He doesn't need to be here. But he's here because what? Because of culture, because of what you're building.

29:25 Canon Reeves How are you able to do that? And how do you maintain that? I think we've hit a really special balance of this company. We are working on this really ambitious long term goal. But there's a lot of short term progress that can be made. And it's technically interesting, but not infeasible. And it's just a hard problem. And we kind of attract smart people who like hard, crazy problems. At the same time, there's this huge upside of if it works, it changes so much.

29:48 Mosheh Poltorak And so I think it's the right intersection of feasible, but difficult and high upside that it just attracts these like crazy, talented people. And then once they're here, how do you keep them motivated? How do you keep yourself motivated with these long time horizons?

30:00 Canon Reeves Yeah, you know, there's so much to do on the day to day. Our team is very small relative to what we've raised and what we're tackling. So there's a lot of work to do. So we don't really have any problems staying busy. We stay kind of focused on like driven towards sort of the next pilot. So Justin's always working on relationships with new customers of different markets, too. I kind of figure out what are we going to do next? And engineering is kind of focused on like, right, here's this next pilot. Here's the engineering we have to do to get there. And then doing that work. So it's pretty easy to stay motivated when there's so much work to do.

30:33 Mosheh Poltorak Yeah, sounds exciting. So aside from the Wendy's pilot, which is massive in its potential and just the name recognition of who you're working with, any other exciting stuff you're working on that you can share?

30:43 Canon Reeves No, none I can announce. We have a few more restaurant partners that we're working with and some exciting, like master plan development and even some citywide developments that we're working with. But, you know, this is one of those.

30:53 Mosheh Poltorak Yeah, I can't I can't share too much. I was hoping you were going to slip up there. Yeah, the stuff that's coming. Awesome. Well, I can't wait to see that. We're nearing the end of our time and we typically close out with a lightning round. Sound good? Yeah, hit me. Let's do it.

31:06 Canon Reeves So what's one book, newsletter and or podcast that you find yourself recommending most often? Acquired. I guess I love to read the books.

31:15 Mosheh Poltorak They recommend they've got great books. Yep. Any specific episodes recently? I'm a big fan of theirs.

31:20 Canon Reeves So anyone that you they liked, especially they did the retail series really well. Walmart, Amazon and they're doing Costco right now or just at Costco. And those are really fascinating.

31:28 Mosheh Poltorak The retail industry is just super fascinating. The Costco one opened my mind in so many ways, even though I like I've been a retail at Home Depot after the company I was with was acquired by them. But that episode was fascinating. I love what they're doing. Yeah, I'm not finished yet. So don't spoil it. No, no, no. Good recommendation there. What about productivity? How do you stay productive?

31:47 Canon Reeves You got a lot on your plate. Any hacks or tips that you swear by? I wish I did have any. You know, I'm an obsessive builder. And so I personally I like to work late at night and just kind of do my CAD mode then. But, you know, I'm not I'm not a huge productivity hacker. I'm more of a just kind of do the work in front of you. And I couldn't not do the work if I wanted to most of the time. OK. What would you change about the startup world? If you could change anything about the venture back startup landscape? Yeah, you know, I think seeing more investment in hardware and people tackling hard problems, we've been really fortunate to find a lot of great investors. But I know that like deep tech and hard tech, I feel really strongly about it being where a lot of the next large generational companies will come from. And so I think just broadening the ecosystem and investments in those sort of plays. And they're a lot different than a software company. It's it's both are hard in their own ways, but hardware has these like really hard challenges. And until you get more funding, it's like there's a lot of hardware companies that don't get to get started. Yeah. Investors are generally wary of hardware. Rightfully so. It's a hard industry.

32:46 Mosheh Poltorak Hardware is hard, right? Harder is so hard. That's what they say. What's a common misconception that people have about you?

32:52 Canon Reeves Hmm. And generally like a CTO, most of my day and most of my obsession is related to the tech. But I was CEO of my last company and did a lot of the sales and I was selling to school. So I was really having to navigate these like large, slow moving government organizations to sell them hardware, just so hard to do. And so I am so passionate about business and product and distribution as something that is upstream of the tech. So really, when I think about building a great product, I think first about the customer and first about the business, the tech. You can always figure out, but those business related things are so key to getting a good product, right? Yeah. If you could have coffee with anybody that are alive, who would that be? It's a great, great question. Sam Walton would be so interesting. Yeah, I love to talk to Sam Walton. What would you talk with him about? I think just retail. I, you know, he's such a fascinating person. I went to college in North Arkansas, so I kind of lived in Walmart world. And, you know, I read his book and, you know, I think he shaped so much about how modern America interacts with physical goods. And like, I think that things are going to continue to change. And I would love to hear his perspective on where stuff is at and where it's going to go. Awesome. What's a core value principle that you live by or try to live by? I think there's just basically no scenarios where total honesty isn't the right move. So I think just being like high integrity, high honesty, men are so much more than money or so much more than like any other thing. So I would say just like high trust, high honesty with your with your team

34:15 Mosheh Poltorak and with everybody, it's super key. Can you share something that you've learned recently unrelated to your work, just something you came across that that interested you?

34:26 Canon Reeves That's a good one. That's a really good question. You know, a lot of the stuff that I love to learn about is just like really niche mechanism design. I follow a ton of Instagram and YouTube channels that are just like creating interesting hardware. There's like this YouTube channel I follow. It's just this like really old mechanical engineering professor that just creates a couple of new CAD linkages and interesting mechanisms a week. And I love watching those. It's all just like, yeah, I learned so much too, as I go through the building process and like I'm always learning new manufacturing methods and new sort of like techniques to achieve certain things.

34:57 Mosheh Poltorak It's going back to your mechanical engineering roots. Yeah, I like it. Cool. Anything else that you'd like to share that we didn't cover?

35:04 Canon Reeves Anything that you wish I would have asked? Nothing that I wish you'd have asked. Only a plug. We're always looking for great people to join. So if you're an engineer or a former founder that would love to do growth or sales, like reach out to us, we're looking for great people to join the team. And would that be remote or in Austin or both? In Austin. You know, we do some remote, but it's only for like really, really, really special candidates in special cases. I think going forward, we kind of just want to do in Austin. Yeah. Yeah. With hardware development, it's hard to do it remotely.

35:30 Mosheh Poltorak Oh, for sure. Yeah. Yeah. You know what? There are companies that you hear about, they get hyped. And then there are companies that you you see that there's something special going on there. And the way that people talk about pipe dream. In fact, there was I forgot who was recently posted on Twitter. You probably know the thread I'm talking about. One of these Twitter celebrities was posting about like, who's doing something super interesting in the hardware space? And I went there quickly to give you guys a plug. And immediately I saw like 10 answers in the pipe dream. Like, I don't I don't need to add anything here. Like everybody is plugging you guys. You guys are doing something special here. And, you know, the world is recognizing that. I can't wait to see what you come up with and where this leads.

36:05 Canon Reeves Yeah. I mean, you know, the support from Twitter has just been amazing. There's so many people in the ecosystem that have been so supportive from the early days when like this didn't make sense. This is a very contrarian thing. And the fun thing about being in the weeds and, you know, working in the business every day is that what the public sees is sort of like six months behind from where we're at. And so I cannot wait for everybody to see what's coming up and where we're at as a business. And I think that we're going to do some really exciting things over the next year. So, yeah, thank you so much for.

36:34 Mosheh Poltorak Absolutely. If we live in a Jetsons future, pipe dreams is going to be a big part of that.

36:38 Canon Reeves I hope so. Yeah. Awesome.

36:39 Mosheh Poltorak Well, thanks so much for your time. I just really appreciate you spending the time today. I know that you are off to one of the test sites right after this. So thank you for squeezing me in here. Really enjoyed this. Absolutely. Thanks. That's a wrap. Thanks so much for listening and joining me on this learning adventure. Hey, if you enjoyed this episode, do me a quick favor and drop a positive review on Apple podcasts or share it with a founder friend. I really appreciate you spreading the word. Tune back in next week, or I'll have the founder of readme.io for a discussion about dev tools and founder burnout. Finally, don't forget to check out growth.co. That's growthwithouttheo.co if you're considering a fractional CMO for your startup. Wishing you rocket ship success in your startup journey. Bye for now.