Welcome to Zipline HQ. You can see we're just about to cross 2 million total deliveries. All of these aircraft flying around that number, do you want to guess what those are?
Yep. But every one of those aircraft is out making deliveries somewhere in the world. Right now, most of the Africa business is slower. It's like middle of the night, Africa time. But if you come in in the morning, you can like barely see that number. There's so many aircraft making deliveries. This whole building, you know, it's been a huge advantage. Sort of similar to, you know, Tesla originally had Fremont, right? They had giant car factory that was right next to engineering for the first several years. We've just been approved by the FAA to now fly one human to 30 aircraft. We'll soon go well in excess of that. Humans aren't really flying the aircraft. You have a human overseeing a large fleet of autonomous vehicles.
My sn that's pretty advanced.
Welcome to Zipline HQ. Um, you can see we're just about to cross 2 million total deliveries. Wow. Yeah. That'll probably happen in the next three or four days. All of these aircraft flying around that number. Do you want to guess what those are?
Yep. So, every one of those aircraft is out making delivery somewhere in the world. Um, right now most of the Africa business is slower. It's like middle of the night Africa time. But if you come in in the morning, you can like barely see that number. There's so many aircraft making deliveries.
Let's actually, speaking of seeing aircraft making deliveries, let's go check out the rock.
What's the rock? There's a rock.
The rock stands for a remote operating command center.
I thought it was actual rock this whole time.
I thought it was like an actual rock this whole time.
You guys are like, 'Let's go see the rock.'
Okay, maybe there's like a fan of rocks.
I mean, we're in California. There's like people who love crystals. Maybe you guys love rocks.
Do we seem like the type to be into energy crystals to you?
We're an engineering company. Molly. It's the opposite of that.
Um, yeah. This is actually kind of a, you know, cool example, by the way. You know, just getting a sense. So, yeah, October 10th, that's what that space looked like where we were just standing. And then December 10th, it was like completely built out manufacturing facility. Um, so this is all of Zipline HQ behind us. Zipline is about 1300 people. We're about um 800 people in the US and the more than half of those people are engineers. Everything from electrical engineering, firmware engineering, all different disciplines of software. So all of those teams are what you see behind us. Most of this column is actually the autonomy team. So responsible for machine learning, uh, detect and avoid, computer vision, both on the zip as well as on the droid. Um and then you know behind them is like uh system uh systems test embedded engineering mechanical engineering supplier industrialization engineering and then over here is
You guys are going to need a bigger office soon.
Yes. In fact maybe you just saw something that we can't announce yet but uh yeah zipline is soon going to be expanding offices. Um this building actually used to be a Harley-Davidson dealership.
Yeah. So they would actually do all of the maintenance of the bikes would happen downstairs and then upstairs they like had all the motorcycles. This was like the showroom.
Zipline. Yeah, it's a lot of tough days trying to build a company like Zipline.
Maggie. Maggie. Zipline chief of staff. Hi. Hey,
We have a lot of questions for you. Are you prepared?
So, this is mostly like powertrain, electrical engineering, more firmware engineering. This is um the safety and reliability team over here. Um and yeah, if you want, we can go check out the rock.
Let's go check out the rock. Yeah.
So yeah, rock stands for remote operating command center. Um, this team is responsible for all flights happening everywhere in the US. We've just been approved by the FAA to now fly one human to 30 aircraft. We'll soon go well in excess of that. But essentially, you know, humans aren't really flying the aircraft. You have a human overseeing a large fleet of autonomous vehicles, if that makes sense. Um, and you can see we track everything from like, you know, uptime to, uh, the kind of cohort analysis of how quickly these sites scale. Um, Zipline's now operating about 20 sites in Dallas alone. Each of those sites scales faster and faster. So, you know, it took us uh it took us, I think, about 110 days to get our first site to 100 deliveries a day. We now have sites that hit 100 deliveries a day in two days. The sites are just scaling way way faster. Um, in some cities we have over 50% of homes ordering from zipline.
It's kind of kind of insane.
Yeah. More than half of all homes ordering from zipline in some of these cities.
Oh my gosh, that's crazy graph.
Yeah, it's pretty cool. Um, we're definitely in the vertical part of the curve. And then if you want to come a little closer, you can actually see these deliveries happening live.
Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three ventureback startups in the US. Nearly 40% of startups fail because they run out of cash. Rex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Rex is designed to help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds. Get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one with same day and even same hour liquidity. Access your funds anytime. Companies like Scale AI, Door Dash, Service Titan, HIMS, Anthropic, Flexport, Robin Hood, and Plaid. Trust and use BRS. Start today at brex.com/sourcy. That's brx.com/scory.
So, you can actually see, you know, there are uh uh there's one aircraft actually making deliveries. This is in P Ridge Bentonville. So, you can see all the different 20 sites are being tracked here. And if we look in, you know, P Ridge, we can see two deliveries happening right now. So there's a zip, you know, in the uh northeast there, you can see a zip that's making delivery. Now it's on its way back. And it also looks like there's a zip that's outbound to make a delivery in the southwest. Um and then as more deliveries come in, um we'll get to see packages get loaded into these aircraft. Um and then the aircraft take off and go make deliveries. So we might we might get to see that over the next over the next two or three minutes. Here you can also see all the other sites and they almost all have deliveries that are happening right now. So
Yeah, that's a lot over there.
Yep. A lot of these sites are I mean, you know, we're right now we're operating sites about um 17 hours a day. We're soon going to transition the 24/7. Um it's kind of insane, you know, just getting to watch like the CCTV. And by the way, the system's growing 15% week over week. So it's, you know, I'll go like 2 weeks without coming and really like watching The Rock. You come back two weeks later, it looks completely different. like the system is scaling so fast almost every two weeks it's like a completely different level of scale that you can observe by watching these flights.
Um so we'll actually get to see here probably in the next 60 seconds you'll see a zip actually hover down and land um on into those docks. So all of this is controlled by this team. So one central control team for all aircraft in the US. Actually sorry there there there are two control centers. This is one of them. Oh, look. A new a new um delivery was just placed. So, see how the see how the um
Bullseye just appeared? Yeah. Now, down in the southeast. So, that means that a zip has just been given a route. We'll probably see a zip take off here. I don't I didn't
Actually see it take off. So, yeah, if you give it maybe 30 seconds, we'll get we're probably going to get to see a zip land here and a zip. Oh, yeah. There you go. See top right? That aircraft's taking off.
There it goes.
So, it's had its route automatically loaded onto it. Now, you can watch it go out and make that delivery um to the south uh to the southeast. There's also a zip that is landing because we just saw it fly back, right? But we haven't seen it land. So, let's watch and we'll see you get to see a zip hover in and land.
Yeah. I mean, coming over and watching this is honestly, you know, it's a little bit mesmerizing because it's like these systems are just now operating.
Feed for the public to work.
Maybe we should. We don't have that. It's like a zoo camera.
Yeah, it's probably a zoo camera.
It's not a bad idea. I wonder if there would be
Put the googly eyes on them. People go. Look, there's a zip coming up after making a delivery.
Um Yeah, we probably should do something like that. It's so fun to watch. Like I would I would watch it just from home.
Um and keep in mind like we're watching the CCTV of one of the sites here where there 20 other sites that are just cycling.
Put the googly eyes on them, start naming them, and then all of a sudden you have a zoo.
Yeah. Yeah, I think it's a good idea.
Yeah, a lot of the uh powertrain teams, so designing everything from battery management systems to the flight computer to the uh power distribution boards. Um and we actually design the electric motors from scratch. In fact, let me we can pick one up in a sec. So, um what you're looking at is HDL. HDL stands for hardware in the loop simulator. You know, Zipline has 140 million commercial autonomous miles with zero safety instance. That's hard to do when you're constantly iterating on all the different aspects of software and hardware in the system. Right now, every six weeks, we're doing a major global software update to the entire fleet. Uh, typically that'll include like 60 bug fixes, 25 30 performance improvements, plus a lot of major new features. All of that has to be really carefully tested before we put it on an airplane, right? And you do not want to like find out for the first time that a feature is failing by putting onto an aircraft and crashing it, even at a test site.
Every one of these layers is you can kind of think of it like the brain and nervous system of an airplane. So this is the battery, battery management system, cellular antennas, Wi-Fi antennas, flight compute, power distribution board, the four electric motors which are designed from scratch by Zipline, plus the red motor controllers you can see on top, winch motor controller, mix motor controller, so the droid brain. This is actually an Nvidia GPU running on that small cute robot that's actually, you know, controlling its own position when it's delivering to a house. plus the dock electronics. So that larger board is what's controlling that big hockey puck dock that's the landing station for the zip. So every one of these layers is an airplane. So and every airplane is connected to a powerful simulator that is simulating a world by that airplane to fly around in. It's kind of like think of them all as like dreaming, right? So you're surrounded by like 30 aircraft right now. They're all flying in the Matrix and we'll download all new software onto this fleet. We can cycle through tens of thousands of flights in simulation running on the actual flight avionics itself. We can typically catch 90 to 95% of bugs and performance software problems here before we then go to the final layer of testing which is the test site itself where we're going to deploy onto real aircraft and fly at scale. Once we do tens of thousands of flights at the test site and everything's working perfectly, then we can do an overtheair global update to the to the to the national fleet of that new software version. kind of makes sense.
Yeah. So, this is all, you know, kind of I mean, you know, it's in the weeds, but um this is a huge part of Zipline's competitive advantage. Like this kind of infrastructure, this kind of investment in testing and validation is how we move at a very high engineering velocity while still deploying a product that's super safe for the neighborhoods that we fly over.
I feel like I'm in a secret laboratory.
Yeah, that's what this is.
Turing is training the next generation of AI with tasks that require real expertise and realworld judgment. That's why companies like Nvidia, Anthropic, Salesforce, and Gemini partner with Turing. Turing builds realistic reinforcement learning environments and data systems based on real operational traces. The kind of infrastructure frontier labs need to train super intelligence. visit touring.coms rce y.
So you're gonna kind of see where we're building all of that. Um this is the current manufacturing line for EV3. So uh primary structures start here and one structure. This is all the structural strength way to
Oh my god, it's really like
Try to pull it apart. You know, you can try pull it apart.
Yeah. Yeah. Try to break it.
