The AR Show: Andrew Hart (Dent Reality) on Connecting the Digital Layer to the Physical World at Scale

5:22PM Apr 11, 2022

Speakers:

Jason McDowall

Andrew Hart

Keywords:

build

ar

great

technology

store

navigation

bit

experience

retailer

solution

thought

interface

point

problem

big

location

imagine

people

find

idea

Welcome to the AR show where I dive deep into augmented reality with a focus on the technology, the use cases and the people behind them. I'm your host Jason McDowall. Today's conversation is with Andrew Hart. Andrew is the co founder and CEO of Dent Reality, a company building the digital layer for the physical world. Andrew and his team have solved key technology problems across indoor mapping an accurate indoor location at scale. And they've created a user interface centered around people, not AR gimmicks.

Andrew is a CEO by day and a pioneering developer by night. Prior to Dent Reality, Andrew pioneered the location based AR space building the most popular open source project for Apple's AR platform, and demonstrated AR navigation and point of interest landmark experiences for the first time, which have since been implemented by both Google and Apple. Today, that reality is working with two of the UK's largest retailers to develop solutions for shoppers and store employees. And reviews dent realities plays to be similar to stripes payment solution,

Stripe realized that lots of people wanted to sell stuff online, it was super difficult and complicated. And for all these financial reasons. And they just came along and said, Look, we've dealt with all of the technical complex problems. And we're just gonna like give you the simple interface. It's a thin layer of technology, which just enables commerce. And lots of people can sell stuff online through that. And so we basically want to be the same thing. For this kind of spatial technology, we can be this one platform, people can sign up, they can get it going really easily. And we can just take care of all of the specialist stuff. And they can build on top of it with everything that they're experts. And that's what we really want to be doing.

In this conversation, we go on to talk about the benefits of building a holistic solution, rather than a series of loosely integrated point solutions, as well as understanding where to draw the line between what you build and others build. We also discuss Andrews approach to building things customers value, and Northstar core values and the transition from developer to CEO. As a reminder, you can find the show notes for this and other episodes at our website thearshow.com. Let's dive in.

Andrew, in the last episode of the podcast, I had a chance to speak with Dana Kamesh, who is deeply passionate about cities and urban development. And you live in or near one of the great cities of the world to in the London area.

Yeah, so I live in Greenwich in London, which is right next to the kind of meridian line and kind of the center of the at least the center of the world in terms of longitude. And I live on the river in I live next to a few interesting things, such as a few museums and stuff like that. And then there's this big kind of old set of buildings, which right now is being used as a university campus, right. It's been a university campus for maybe the last 50 years or however long. But for hundreds of years before that it was like built by Henry the Eighth and used by him and in over time, it's evolved. And so it's not a university campus, but it's super old and super grandiose. And so they use it in all sorts of film productions. So to give you an idea of the type of building is, if you've ever watched the crown, you've seen Buckingham Palace, that that's that's what that is. So right now, they've kind of transformed this university campus into this giant film set for new film Colin Polian directed by Ridley Scott. And so I walked in the other day, they've got like, all of these horses and all of these people dressed up in ancient kind of 1800s esque clothing, and all of these giant lights trying to imitate being the moon and stuff like that. Giant cranes and camera crews and all of these stuff. And so they've literally got about 400 people there. And so yeah, they're kind of firing cannons at each other and everything. So if you do hear any, like cannon fire or anything in the background during this interview, that's that's the reason why

that's pretty amazing. You get to live in multiple cities at the same time sometimes, right? And multiple time periods.

Yeah, literally, you walk in, and they've got these massive French flags up and everything like that. And everything is decked out to look like the 1800s like or I guess it was like 1800 to 18 or five sort of time period. Yeah, pretty crazy. And I've also like, that used to be my running look when I used to when I used to run, I used to be a bit more healthy than I am now. And I remember this one time, I had my Apple watch, and I was timing my run. And I was running all the way around and I came back and just like within the last kind of half a kilometer of my run before I got got back home. I was running through the campus. And there was Matt Smith, and it was Elizabeth Foy filming the first season of the crown. And so they were there and they're, you know, kind of ancient or 1950s sort of cars and stuff like that. And they were just right there. And I just ran straight past them as I was doing my run and they were in between tech so it was okay. But I was like do I just stop my run and lose my record because I was like on a on a streak here. And I was like, but what I did was I just ran all the way I finished my run, got my record and then run back and was like, just watch them filming. So

that's pretty awesome. It's pretty funny. It'd been even more funny. Had they been in the middle of filming as you sprinted through the back of the scene with some people and equanimity here. Yeah, awesome. So this tie of the city, this notion of place is something that you are also very passionate about. But for you, your background is in software development, software engineering. And you are really focused on tying digital objects to the real world. When did you first fall in love with that sort of concept?

Yeah, so this is really interesting. The first time I ever worked with location technology, was when I was working for a startup in Pittsburgh. And we were working on a game, which at the time, there was lots of lessons from the startup, one of them was that we didn't really, we didn't not do a good job of user testing, and finding product market fit and all of that kind of thing. And this was in my early 20s. So that's fine. But one of the lessons that came out of it was that the idea of what we were building was not bad. In fact, it went on to become one of the most successful games in the world. Because it was a, it was pull Come on, but played in the real world with real world location. And so we did that. And we released it. And, you know, nobody downloaded it, the product itself wasn't wasn't where it could have been if we maybe spent more time on doing rapid testing and all of those things. But then, you know, cut to a year or two later, and then PokemonGo comes out, and give us some real validation that actually, you know, what, we were onto something with the idea at least,

did you actually have Nintendo rights to Pokemon? Or whoever owned the Pokemon Company?

No, not at all. I think there were two issues with what we were doing. Number one was we didn't have the rights to it. And so we didn't, we didn't infringe it, we had these awesome artists and designers who came up with our entire new cast of characters that you would play with. And the second thing was about the product market fit. And I think, if the product market fit thing had worked, I think the the fact that we didn't have the Pokemon trademark, or those characters, I think that side of it would have been fine. But there was also that aspect of it of like, was it the branding of Pokemon that helped that to become so popular? Yeah, a big part of it? Yeah, amazing. Actually, there's an interesting story to tell about that game. So it was called megabits. And this was really my first if you talk about augmented in the real world, this was kind of my first foray into that, we actually weren't using augmented reality at all, that was one of the big differences. So I didn't have any any kind of AR experience from that from building that. But because this was like 2014, or something like that. So one of the things that we did was, we knew that we wanted to play it on top of the real world map. But what we didn't want to do is we didn't want to overlay it on, you know, Google Maps, Street View, or sorry, Google Maps, or satellite maps, or, or even just the regular street maps, because we wanted it to be gamified. And so we said, okay, which direction do we go with this? Now, if you've played Pokemon Go, you know that they just apply to kind of cartoony effect to the, to the void layout and everything, they just apply, like sort of a custom a custom skin to it, we thought we would take a completely different approach to that. Of course, nobody had done this before. But we we kind of thought about a few ideas. And we thought, Well, what do we want it to look like? We want it to look like Pokemon, right? We want it to look like those types of games, where you have that kind of 2d grid. And if you've got a pathway, it kind of walk goes off, you know, straight and then it goes up, and then it goes along. And you've got houses also on this kind of 2d grid. And you know, any kind of water or grass or anything like that. And so we thought, that's really what we want to do. Now, the problem was, is that the real world is not a squared off grid, right? Like roads go in all sorts of directions. And so what we did was we came up with this really clever system, where we would just basically took the satellite imagery of the entire world, and we would look at each or maybe maybe it wasn't satellite imagery, maybe it was just the regular kind of Apple Maps, you rendering of each tile, each map tile, and we took it and we would look at it, we would analyze the the image of it, the colors of it and everything. And we would say, is this a road? Is this grass? Is this water? Is this a park? And we would do that, and then we would put the title on the map relevant to what it was. And so what we ended up with was this very old school designed map, which will look like a kind of a game, but it was the real world map. And that was one of the coolest things that I think we built from that experience.

That sounds super interesting. So you're basically analyzing you're interpreting through computer vision. The LLS

computer vision would have been nice, wouldn't it? Now back in those days, when I was developing that part of it, that was probably about 2013. So computer vision and deep learning just was not at that point. If I was I was saying to the guy who was CEO of that company quite recently, you know, if we were doing that today, we would 100% do it with computer vision. And we were just, you know, pumping loads of images tagged and everything at the time that did exist. So this is like a perfect use case for computer vision, where we manually looked at each of the different things around, okay, this is white, therefore it is void. This is this color, therefore to this, and then we have the challenge of well, you know, on those map tiles, the text is written in white with a great surrounding to it. And so we would go okay, well, this pixel is Wait, now let's extrapolate outwards. Can we see any gray? No? Okay, great. It's a road. It's not any text. So we have to apply all of this kind of manual logic. And, you know, that's the exact thing that people say all this, you know, what's a great example of machine learning or computer vision? That is a perfect example of where that technology, you know, would have made that a lot easier.

That's pretty awesome. And so from there, so these are the early days, this is pre AR kit, even right? This is before we have this AR capabilities on the smartphone, certainly in the package library, like iOS had made available around that time. Yeah. How did you imagine or how did you begin to take advantage of, of some of that technology.

So when AR kit came out, I was really interested in, there was loads of people doing demos online and showing what you could do with it. And it was really cool. You remember, like when Craig Federighi went on stage at Apple, and he said, Look, a lot of people have shown some videos, which you know, are just very carefully edited. But like, let me show you live something which actually works. And he did that demo on a table with like, I think there was like some video game and some like machine gun artillery or something, you know, everyone was blown away by it. And people started doing demos of different use cases. Now a lot of it was more in the kind of novelty area, the sort of things were reminded me of when people were making the first iPhone apps. And they were all kind of thought ups and in nothing really useful. But Obeah reps, remember that be a thing where you would drink on the phone, and it would use the physics. So people were making those types of things, right, those kinds of demos of oh, great look, look at this new technology. What I thought was really interesting about it was that this was a technology that could go beyond the screen, and integrate with the real world. And what I mean by that is that I kind of see that everyone's interaction with a with technology right now is buried within a screen. It's an interaction with a device. And I thought, what if you could take that interaction? And you could kind of overlay those interfaces on the real world? How would things change if it wasn't like a device with technology packed into it, surrounded by a kind of analog real world, and it became a real world filled with digital interaction. And so that was kind of the spark that made me think this could be really interesting. If it had location functionality, if you could start overlaying stuff, which had real world context. So that's kind of what I started to explore

in how did you end up exploring that? I think, if I recall, you built quite a name for yourself among the developer community at that time.

Yeah, so I didn't especially intend to do that. But I thought it was an interesting idea. I thought it had a lot of potential to it. And so I thought, well, maybe I could contribute something to this by building something that would enable developers to go and build many, many different ideas that you could build with location based AR technology. And so as I say, I live in Greenwich. And in one, it's funny, because one of the challenges is okay, how do you know what your real world location is? How do you know like, your precise location, how you gonna measure that versus what the phone tells you? Well, luckily, living in Greenwich next to you know, this is what Greenwich Mean Time, is named after, there's literally a line down the middle of the road about five minutes from where I live, which shows you like the true north line, and this is exactly on the meridian. And if you stand on one side of the line, then you're on, you know, the negative longitude. And if you stand on the other side, you're on the positive longitude. And so I use that as a helpful anchor in the real world five minutes from from my apartment, you know, to tell me build some of the stuff. But yeah, so basically, I spent sort of nights and weekends working on how to enable location based applications. I basically merged together a series of technologies around Apple's colocation framework on iOS and AR kit, and solved a few kind of algorithmic problems to make that happen. And then I thought, great, well, I've got something I want to open source it I want people to find use from it. So I'm going to make my own demos. And so I the first one I did was like a demo of a skyline, where I showed, here's some points of interest on the skyline, Canary Wharf in London and the auto arena. And I think the shard, which is the huge building here, and so I thought, Great, that's that's decent video, and I put it out on Twitter, I closed it started my workday, didn't think anything of it. And then I come back to Twitter a bit later on, and it's blown up, and people are going like crazy. Like, how on earth have you done this? This is awesome. This is what I want er to be for all of that kind of thing. So I thought, Okay, that was pretty cool. And then I was, I was thinking, Okay, well, I had in mind this idea for a second demo, which was about navigation. Now that demo will take a bit long because it involves kind of integrating with MapKit. And figuring out how to draw the polyline from the from the, that it gives you for a route to a destination, in figuring out how to render that as a 3d object, but I thought, You know what loads of people have reacted to that first video. So let me do another demo of navigation, how you can use the technology I've built for navigation. So I went ahead and I did that demo. And I posted that one online. And that just took off to a complete next level. And people were at that point, we're just kind of going and seeing my notifications were blowing up all of that kind of thing. And then eventually, about a week after that, I open sourced it actually a little bit. The funny thing was the day before I open sourced it, I got this email from GitHub. And this gives a sense to kind of the pandemonium around it by this point, this had been featured on like Mashable and 95, Mark and a bunch of other news websites, I get this email from GitHub saying an issue has been closed on your AR kit repo. And I was like, okay, that's weird. By issue, it means discussion has been closed. And I was like, but I did create a repo, but I haven't uploaded any code to it yet. It's not ready to upload. And I haven't linked anybody to it. So we're talking about. And so I followed the link. And I just found this very, very long discussion thread of loads of people arguing back and forth over like, oh, where's the code? Why he's Why has he not uploaded it yet? Oh, my God, was it all fake? Is he lying? And people like, you're gonna be lucky if Andrew ever shares his code with you? I'm like, This is insane. And then, yeah, so this was this was like some next level of hype around code that I had written, which was just seemed insane. But then I did open source it about a day later, just if for no other reason, then to quell everyone's kind of frantic need to guess. Yeah, it did tremendously. Well, it it kind of rocketed me from being like this kind of anonymous developer who like really didn't have an audience at all, to suddenly being like, top trending iOS developer on GitHub for several months at that point. So yep, so that was my start in AR very, very much a rocket ship. Yeah, that was a pretty crazy time.

How did you deal with all that attention?

I don't know, it was pretty crazy. I just kind of tried to, I tried to encapsulate it and respond to it. I built a little slack community, which was easy enough to do because there was so many people responding, I didn't really need to put any work into it. But I kind of thought I was getting loads and loads of messages with people asking me questions on how it worked and stuff like that. And then a lot of people were asking if they could contribute to make improvements to the code? Surely it needed a lot of improvement. Certainly, I would look back and say that now. And so I thought, Well, what I could do is I could create like a Slack community, and just say, here's the link, and people can join it. And then they can all work together, and discuss together and answer each other's questions and stuff. So I did that. And that kind of shifted a lot of the focus for me, and I didn't need to so much be super involved and feeling bad, because I haven't replied to people's messages and stuff like that. But then I just kind of thought about, okay, what do I do next? Right? How do I take it to the next level? I quit my job by this point. So I was like, great, like, what am I going to do next in the AR space, because there's surely a big opportunity here. But I hadn't quite identified what it was going to be

in, what was that process for you of figuring out what the next opportunity is going to be?

I went in, I just kind of start creating other other demos of different use cases, to just try an experiment, to learn a bit about how to build a special interface, which I realized was very different from building like a regular phone interface, you have to really think about how you're displaying elements and objects in that AR environment in a way that makes sense on mobile. And so I was exploring some of those things, and how different applications could work well. So I think I did a demo in a bookstore. And I did one navigating around an office campus. And I did one in a shopping center, and one in a retail store. And I did a few of the things like that. And these were just proof of concept things for the most part, like most of most of the time, there was like a bare level of technology under them. But these were most mostly kind of let's, you know, post these out there and kind of see, you know, see, see what lines with people and see what, what what comes back from it.

And as you're going through this, how are you actually doing the testing. So you build these proof of concepts, and then where you're recruiting friends, so you're recruiting passer bys or, or corporate entities.

I wasn't really testing it with anyone. It was more about trying to understand what people's reaction to it was. So I could surely like for example, there was like a book recognition demo where you could hover your phone in front of a book and it would kind of give you the information about it. I'm sure. There's been loads of people who've done that I probably wasn't the first one to do it. And lots of people have made better versions of that sense. But it's just about kind of bringing to life, you know, inanimate objects, basically with information and utility. And so it was about understanding what Which of these things do people respond to most important to the find most valuable about them. And then a lot of the a lot of the response to some of those things was actually from big organizations. Like, for example, when I, when I did the demo in a in a grocery store, I suddenly got insane levels of inbound from big grocery retailers. And they were kind of saying, you know, we've been looking for a way to kind of digitally connect with our customers, for all for a variety of different reasons. And it looks like you've built some technology, which does that, like, Could we talk about it? And so I used a lot of those conversations early on to just learn a bit more about like, what are they trying to solve? And what were the challenges that they were seeing? And why have they not been able to solve it before. And that really kind of led into what dent reality has become in me figuring out what the opportunity there was to kind of, you know, build really build a platform around that kind of technology.

So you took that and internalize that feedback that you're getting, and that is what created the foundation for dent reality where you're currently focused. Yep. And within this, you have really focused on pushing the state of the art as relates to indoor positioning, because this notion around navigation, and putting people in the right place with the digital information in the right place. Yep, within a space, what maybe we can kind of take a step back, and you could explain why it is that indoor positioning is so challenging.

Yeah, well, this is a this is a market where it's really funny. We're in an unusual place, as a startup, because we get a lot of inbound, like we got way more inbound than you could imagine. Like I see at least one email coming in every day. That I think, yep, that could be like a million dollar opportunity on its own, we just get so much of that stuff, where I'm like, these are all really exciting opportunities. And when we kind of jump on calls, and we speak to these people, they all basically have the same story, which is that we are not the first people they've reached out to, quite often they have went with one of the provider, or they've tried two or three other people for positioning technology. And they've been sold these really expensive projects to do positioning, and they've invested a lot of time and resource and money in it. And it ultimately hasn't worked out for them. And so they end up coming to us and saying, look, we've been trying to solve this problem for ages. And so like when we heard about this, and when we started hearing the same thing over and over. That's exactly the question we asked, like, why is this so challenging? Why? Why are we constantly hearing about this. And the main culprit is this kind of iBeacon technology, which iBeacons comes from Apple actually, but even Apple says do not use this for positioning, there are better solutions out there. The eye beacons are pretty bad, because they mean that you need to install hundreds, if not 1000s of these beacons around the location, they're very expensive, they are very time consuming to set up, the batteries run out every six months. So you need to get your stepladder back out and run around replacing all their batteries. But even more than that, that may all be fine if they actually worked. But what we hear back and what we've experienced in our testing, is that they typically get the best case about five meters of accuracy. And so like if you think about that, on an indoor scale, you want to precisely understand where somebody is, so that you can serve them up information, whether that's about what they're looking out, or whether it's navigation, whatever, five meters, if you think about a grocery store, for example, could be the difference of two or three aisles away from where you actually are. And it tells you that you're in one place one second, and then a second later, it jumps you three miles away to somewhere else. And then a second later, it jumps you back to the first place. And so it's just this crazy nonsense. And that type tends to be the case for a lot of these other solutions that you see, you see things like magnetometers and using the kind of the magnetic field of the Earth, but it's very kind of inconsistent. So like if somebody so much has opens a door, it throws off the signal and it gets confused. You get computer vision solutions, which work well outdoors, but don't work as well indoors for various different reasons. And so we basically looked at all of that and said, all of these things suck for different reasons, most of all, is that they're not accurate enough. But even above that is that they're not very scalable, because you need to install all of these different sensors and everything like that. And so our approach to it from the beginning was like, Okay, great. Can we use the technology that we understand really well augmented reality? Can we use that as part of a solution where we can get something which is way more scalable, and way more precise than what exists out there? So that's basically what we've been building for the past several years.

You talked about the scalability challenge, and there is an approach to positioning that we've heard a little bit about over the years, which has a similar advantage and scalability, which is using Wi Fi for positioning. Yep, and the idea is basically to triangle Based on a handful of Wi Fi routers, yep. Why is it that these other companies aren't able to solve this problem using Wi Fi signaling, or Wi Fi triangulation?

Yeah, Wi Fi is better than beacons, because it doesn't require you installed depending on the approach you take, you can get some kind of specialist Wi Fi hardware that you can buy, I would not recommend anybody does this. Because yeah, again, you hear the same level of accuracy, about five meters of accuracy, except you've just shelled out loads of money on some specialist Wi Fi hardware. That is also the some latency involved there. So the way that those systems work is, it's about back and forth communication with a Wi Fi thing. So if you can imagine your device and you ping the Wi Fi, and you say, hey, look, I want to know my location of where I am. It, you know, you send a bit of data over and then it propagates and then it comes back to you with an answer. By that point, you've already moved forward. So you're already losing accuracy, by the fact that you've walked away from where you were in the first place when it was trying to process where you were. So there's big latency issue there. And also a privacy issue as well, obviously, because you're just sent, you know, seriously sending out your location, and communicating outside your device. So I would recommend against those solutions. But Wi Fi, if you use the existing Wi Fi, there's, for example, the Apple indoor maps program, which is which has now built on this and allows you to enable Wi Fi positioning in your space. This, again, gets a similar level of accuracy to beacons. So it's better because it doesn't require any new hardware. But it still gives you a result, which is about the same level of accuracy as because now what we've been able to do is we've been able to use everything we understand about augmented reality, what augmented reality does is it knows your it knows the precise movement of your device, right? Like there's kind of two elements of it. If somebody doesn't know anything about AI, kind of tell them two things, which is like, number one is that it lets you it knows the precise movement of your device. So if I'm pointing my phone at my laptop, it's using the process of slum in order to using the camera using the gyroscope using the accelerometer to understand the precise motion of my phone, as I'm kind of moving it around. So if I'm pointing it at my computer, it can recognize my computer and it can, you know, see how the computer is moving in relation to my phone. So it knows your precise movement. And so what we've been able to do is we've been able to take the advantage of over understanding that. And we've basically written up these really deep algorithms over the past several years, which combine AR with Wi Fi. And by doing that, we've been able to use the kind of estimated location that you get from the Wi Fi, combined with the precise motion from augmented reality. And by doing that, we're able to hone in very precisely on where the person is within a within a retail store or any other kind of indoor space. And that just enables you to go to a next level at once you're kind of under about a meter of accuracy. That just allows you to unlock a whole range of new experiences and in applications that you can build.

Let's dive in to this this area, this application, this sort of experiences unlocked with this sort of capability. You had talked about how kind of the genesis for dent reality came from all the feedback you're getting from the grocery retailers. Yeah, about the demands that they had the needs that they had, how are you applying this technology to fit those needs?

Yeah, well, it's, it's interesting, there's kind of two use cases, within grocery retail, we decided we would start with that, because we had so much inbound from them. We also get a lot of inbound from many, many other markets, but grocery retail came to us. And they have two distinctive needs, right? Number one is about the customer experience. There was an article written recently, I think it was on the hustle about all of the all of the different techniques that IKEA uses, in order to boost the customer experience in order to get people to buy more stuff in order to, you know, improve their stores, basically. And they harness every single trick that they can, in every single technique to really drive sales. And, you know, so meticulous and they've invested a lot of work and energy into making that happen. And so if you can think Imagine if you could give them a tool that could understand precisely where one of their shoppers is in the store, and then deliver them an experience. Imagine what they would do with that technology. Imagine all of the ways that they would just go great, fantastic. We're going to harness this and we're going to go wild. And we're going to take everything that we're doing already, where we know nothing about each individual customer. And we're just going to take it to the next level. So what we want to do is we want to basically provide a technology platform, which just we can we can solve all of the specialists stuff that we understand about such as precisely locating people building great spatial interface, enabling them to map their locations super quickly. And then they can build all of the value that they want to build on top. Right, I use IKEA as an example. But you can think the same in grocery retail, right? Everyone has heard all of those stories about the different ways that grocery retailers organize the store, and the layout and everything in order to maximize value of shoppers and stuff like that. So you can imagine all of the ideas they might have, about how they can apply that using this technology. So that's kind of the first thing is on the consumer side. The second side, which is really interesting, has been an uptick since COVID, really is about it's about pickers in the food delivery side of things, where they are seeing an ever rising increase in the number of people ordering food for delivery the next day. And they've got these kinds of pickers running around with these terrible devices in the middle of the night. And they're given really bad instructions. So it could be something like, Hey, look, you know, this item is on aisle 13 On the left, but Okay, great. What does that mean? Right, on the left is, you know, it depends on which way you enter the aisle from if you enter it from the side of the aisle, then left is on the opposite side. So the tools that they have for this are really bad. And they just are looking at a device at three o'clock in the morning, which has a bunch of code words on it. So it's really terrible software. And so what they've kind of come to us and said is, look, we'd love to use your technology as a way to guide these pickers around so that they can deliver more food quickly, basically. And so those are the two really big use cases that we're seeing in grocery retail that we're starting to explore.

Amazing. So in the picker side, on the on the enterprise side, commercial side, they're really the value delivering is time to completion, I the the rights of the employees, or the pickers have the right set of guidance, much more precise, and they can get through each one of those orders that much more quickly. Yep. And on the consumer side, the opportunity is to from, from the enterprise perspective, the opportunities to increase the amount of perhaps spend that they have in the store. But from the consumer side is surely about enhancing the experience and getting through that shopping more effectively. More delightfully, maybe even as well.

Yep. Yeah, I think that's a big part of it. You know, if you think about when your current experience of going into a grocery store, this is kind of the analog world that we live in, is that you've got all of these kinds of red, special offer labels everywhere in the store, and you ignore all of them, you just your brain filters them out, because there's so much noise. So many of these going on that you just It's overwhelming, and you ignore all of them. And the reason that stores do that is simply because they don't know who each shopper is, they don't know what's interesting to every shopper. And so they can't just tell you that and they can't say well, look, we're going to tell you about the one or two items that we think you should really know about, for example, like, hey, look, this item that you usually buy, it's on a special offer, right now, you might want to buy it and add it to your basket, that will be a really useful thing for them to be able to communicate to a shopper. But they just have no way of doing that right now, they can do it very effectively online, through online shopping, right, they can say here are some recommendations, every single person's recommendations have different, they have no way of doing that and delivering that in store. So it's, there's a side of like, you're able to increase the basket value for the retailer, you're also able to increase that customer experience at the same time. And I think those two things are really important because one of the beautiful things about it is that you need the consumer buy in, right? If you if you invent the solution, and a retailer goes great, we're gonna bombard people with annoying stuff that they don't care about, great, well, nobody's going to use it. So you're kind of incentivized to build something that is an awesome customer experience. Because then people will be like, brilliant, I get loads of value out of this, it shows me the offers that I'm interested in, it helps me find everything on my shopping list, I don't feel lost, I have all of the guidance that I need in the store, I can get through it really quickly. And then the retailer wins, because they're able to, you know, sell all of the relevant stuff to the customers. So

you know, a store that I desperately wish had something like this is Costco, here in the US is kind of the the bulk discount sort of store you go in and you you buy more of an item than you typically need. But there's kind of more of a warehouse sort of shopping experience. But I find that whenever I go to the store, they always move the things around. Sometimes they don't have the items anymore, or they don't have them currently, or they've moved it from one row to another part of the store. And I feel like we spend always twice as much time as we need in the store just trying to track down the things on our shopping list. It's pretty painful. So I look forward to this being in the US as well. Yeah, absolutely. One of the things you talked about earlier was one of the challenges around the user interface or on the user interface. This Yes, this notion of kind of building a spatial interface. What have been the biggest challenges for spatial interface design?

Yeah, there were lots of areas of augmented reality that get a lot of attention and volume of attention. I think the design of the special interface is one that just I barely see anybody ever talking about, but it's it's one of the most critical things that you You need to get right is how do you build a compelling experience? For example, the very first way that you could go about it is you could give someone an experience. Let's take navigation as an example, where somebody has to hold the phone up and with the phone in front of them the entire time, right? In order to see anything. Well, that's uncomfortable, it's uncomfortable, physically, you don't want to hold your hand up the whole time. And it's also socially uncomfortable, because you don't want people to think that you're filming them or anything else like that. Right? So we've always believed that that's the complete wrong way to think about it. What you want to do is we want to build an experience for mobile, because wearables are different, right? wearables, the paradigm shifts a little bit, but you want to think about how do you build a great mobile AI experience, that just works in a very natural way. And so what we've come up with we've we've done a lot of research and development on this over time, in terms of how you display navigation elements, a lot of people start with the path on the ground, I started with the path on the ground. And then we figured out actually that isn't the best solution. We think about it in terms of the way that you might scale information, your initial approach might be to show a virtual label, like a virtual text label, and you show it maybe at the end of the aisle, and then you realize these re squinting at the screen because it's too small, and they can't read it. Or maybe they get close to it. And now it's huge, and they have to pan their phone around to fit the whole thing in. So there's a bit of thought as to Okay, well, how do you make sure that that AR content is both scaled in, in an appropriate way, where if you if, by looking at it, you can see how far away it is? Because you're familiar with how big those elements are supposed to be? But at the same time, it's scaling so that you can read them, right. This is an interesting kind of conundrum. Is that how you solve that problem? And one of the third things is about is back to that comfort thing? How do you integrate that naturally, with a regular kind of UI for fun? Let me talk about the third point. And if you want, I can talk about what our approach has been to those other two points, what we've built as we built this incredible interface, which, as soon as you try, every single person, I put this in the hands of has this aha moment, it's a bit like pull to refresh on the phone, where what we've done is we've got a full screen map or full screen shopping list. We've got a compass mode, all of these things that we've we've built in tried. And depending on which version of the app you're using, we switch them in and out. And we're continuing to do our user testing. But we've got a regular UI that you can interact with. And then you hold the phone upwards, right, because you're using it in a normal way in your hand. And then you hold your phone upwards. And then the map interface just drops down. And you see the camera mode, right, you see the augmented reality. And you can see all of the instructions. And you can see the screen edge indicator, like the thing that you get in video games telling you, you need to turn around to your destination. And so you can see all of this, and it's very natural, you pull your phone back down, again, to a more natural place. And the interface brings back back up the maps and the shopping list and all of those other things. And so you can switch between those things very naturally. Recently, we went all in on a compass design, where you've just got a full screen compass. And so you hold the compass in your hand, and you can, you know, move the phone around. And you can see that it is very precise. And it moves with the motion of your hand. And it shows you exactly which way to go. It shows you how far you need to go. It gives you all of that information. But you bring the phone upwards and the compass actually transforms into the AR view. So now the same information is projected in front of you, on top of the world, you bring the phone back down again, it's back in the compass. You know, this was not an overnight discovery, like it took a long time of research and development to discover how to do this. But it's absolutely essential. If you want people to use your technology, basically,

that's deeply insightful, a lot of back and forth to kind of figure out the right level of friction, the right sort of value in the right places. So that that type of interaction is pretty awesome. Let's go back to the the first one you had mentioned and touch on that one, at least, you talked about how everybody thinks that we need to put, you know, yellow brick road or whatever the equivalent is the line. What have you found works better than that?

Yeah. So it's interesting, from everyone's lived experience of using Google Maps and stuff like that, you've got kind of that blue line across across the road showing you air to be navigation. That's always everyone's first approach whenever they, you know, arrive in AR and they try and do navigation. It was my first approach as well. I did that for a while, and then discovered that for various reasons. It didn't work very well. There's lots of really, really important but completely distinctive reasons for why that doesn't work. Let me tell you about two of them. The first reason is that sometimes in AR You can launch into the AR experience and you have that thing where you need to wave the phone round in order for it to calibrate, and understand your surroundings, the reason for that is because what it's really doing is it's trying to understand where the floor is how low the floor level is, because the experience that you're going into requires placing something on the ground. And so it's trying to scan to understand where the floor plan is, right. And so you have to wave your phone around and everything. And you can imagine that's just a friction point, it's a barrier to entry, if you want somebody to use your experience, but it's required if you're going to put anything on the ground. And so what we realized was, look, if we just raise all of these elements up into the air, and have them float in front of you, then we can save the use of that entire setup process. And you can just instantly get into AR. So the second that you open it, you can hold your phone upwards. And you can see those elements right in front of you. And there's no calibration step required. And you also save about 10% on performance as well, that

totally makes sense, right? That there is this physical fiction, there is this performance issue that comes with figuring out where the ground is upon which you will then lay the blue line. So that that probably makes sense. What what was one of the other ones,

I'll tell you the second reason, this one is a little bit more in depth to explain. But hopefully, hopefully, I can explain it in a clearer way. Without diagrams, imagine that you're just laying out the path on the ground in front of the person, there's a few things that need to go right for this to happen, you need to understand precisely where the phone is great, we've got technology, which can locate somebody to enter a meter of accuracy. Checkmark, you need to understand the precise heading of the phone. Now the compass on the phone is about 15 degrees off at the best of times. Now imagine that you're laying down on the ground for someone, and that's off by 15 degrees, you know, it's going in the wrong direction, that looks terrible. So you need to solve that problem. Now thankfully, our technology has overcome that. And we have about one or two degrees of accuracy on heading, which is incredible, by the way, that's kind of beyond the level by which you then can't distinguish. If it's correct or not like the human eye looking at one or two degrees, it looks correct. So it's like checkmark, we've made that we've made that leap. So we've overcome those challenges. But then you lay the blue path down on the ground for navigation, right. And if you imagine a path going in front of you, and then turning left through a doorway, and imagine that your accuracy of that experience is about a meter, right. So let's imagine that the the paths going through the door is off by a meter, right? You're walking along and you see that path, even if it's off by half a meter, let's say you see that path turning through the doorway, you can now look at that path. And you can see oh, it should be going through the middle of the doorway, I can see that it's going through the edge of the doorway. And so without even being a scrutiny person without being like a, you know, a super detail person, you just look at it and it looks bad. It's almost like by drawing that path on the ground, you have drawn the template of look, this is exactly where we think you are. And somebody using that is going to look at that. And they're gonna think, you know, it's not accurate, like, oh, you know, it just point it's, it's like drawing a template on the ground of how inaccurate you are. It's showing people it's almost like a debug mode of like, this is exactly like where we think this entranceway is. And you can look at that. And without much scrutiny. You could say, oh, that's clearly wrong. And it's a shame because it takes you out of the experience. And it makes you focus on the fact that it's a bit wrong, more than the fact that how incredible it is that it's that accurate. And so what we realized was that, what we could do instead of that was that we could use basically a system of hovering arrows and what became guidance dots, where we show all of the dots in front of you up to this big hovering arrow next to the doorway, if you can imagine this. So you walk along, and you're just walking through and these dots are kind of in front of you. And they're disappearing as you approach them and everything. And you walk up to this big oven that I've always hovering right there in the doorway. Now looking at that level, that level might be a meter behind the doorway, and might be a meter in front of the doorway, or it might be right in the middle. You don't know whether I was supposed to be. And so if you're looking at it as a regular person, you're like, Okay, great. It's in the right place. It's pointing through the doorway. Great. I will walk through the doorway. And so what we found in our testing was just a massive improvement, where rather than focusing on precision of accuracy and those types of things, which were just irrelevant to focus on. Now people just use it and it feels like magic. Now we did implement a few additional things to smooth that experience and make sure that it will be entirely elegant and all of those things. This is basically the the result of a lot of research and development right? To figure out what works best and what doesn't work. Were the first ones to do this. But I think if you look a few years down the lineup, people implementing navigation style experiences, for many use cases that we're not going to solve ourselves such as DOS, I think you'll probably hopefully see other people arriving at the same conclusion. So Pac Man had a right all along. Right, exactly. Yeah. The Pac Man, the Pac Man experience.

That makes perfect sense that, although not immediately obvious, this notion that laying on the ground, even if you are 99%, correct, that 1% Incorrect is where people focus all their energy. And so they lose. Secondly, they lose trust, and they lose focus on the value that's there. Yeah,

yeah, exactly. And that's no fault of the person using it, right. These are not people using it to try and scrutinize the technology. They're just like avid users, and they're seeing this inaccuracy. And it's giving them a worse experience, the fact that you're visibly showing them this. So yeah, I think this is the value of, of building the interface at the same time as you're building the technology, like you understand how the technology works. And so they then you can build an interface in step with that at the same time that works well for that technology. At the same time, you can build the technology to serve the interface Well, while you're like, great this is I need a really smooth experience, where we're going to ultimately display this to the user, like give me this and you can you can build those things. So I think like if we were just building one part of the widget, right? If we were just an indoor positioning company, and somebody else was providing the maps and somebody else was providing the navigation, somebody else is providing the the the user experience, there's no way that you could plug all of those solutions together and come up with an elegant experience, you kind of need to be doing all of these different parts of this of the solution in order to make an elegant user experience.

You talked previously about this the challenge around scale? Yes. And you are solving that by being able to leverage the in place infrastructure that the stores already have, which was existing Wi Fi infrastructure, and the in place infrastructure that people already have with them, which is their smartphone user brings part of the solution, and the store has a little bit of the other solution. They don't have to install a bunch of these beacons that they have to replace the batteries on every so often. They don't have to spell any fancy additional Wi Fi equipment. Yeah. But the other piece that you just hinted at there, which is part of the solution offering is the map itself. How do you solve for mapping and doing that at scale as well?

Yeah, well, I alluded to it a second ago. But, you know, I recently spoke to the guy who was in charge of building out a solution like this for the Mall of Dubai, right, the biggest shopping center in the world. And he said, You know, I went to one company, for beacon technology. And I went to one company for maps and navigation. And I went to one company for user experience. And when the maps broke down, and the navigation, and the beacon stopped working, they would all point fingers at each other, and it was a complete disaster. So that's one problem is you've got all of these competing parties in a lot of this frustrating technology, that one of the ways that you achieve scalability with this is you have to kind of do the whole thing as one big piece, right? You have to be the one company that's going to solve this problem, and provide the maps, the positioning technology, and the user interface. And if you bundle all of these things together, you can solve the entire problem. And you can do it in a very scalable way if you have the right approach to it. So a lot of the kind of players in the industry are making their money right now on charging these really big upfront fees. Like they'll say, you know, pay me $100,000 to come in install beacons in one location. And it's going to take months of setup, and it's going to be, you know, they're going to have these experts working on specialist software to map it all. And the approach that we've taken is entirely different. We want it to be as easy and as scalable as possible. And the reason for that is because we want to build the one platform plugged into every indoor space, right? The only way you get there is not by selling your time, but by the by building something where you're not selling your time. And you can you can wake up one day, and you've got a new 100 new stores that you don't know the name of or you don't even know where they are, and they've all gone live, and it's all fine. So that's where we want to get to one of the big challenges with scalability in the real world is that, you know, let's say you you're the maker of I don't know, Wordle, right. And you think 10,000 People are going to use your game one day, and you wake up and a million people want to use it. Well, you can turn up your AWS spend, and suddenly, you know, a million people can access it. Now, if we're in 10 stores with a retailer, for example, and they want to take that to 1000 stores. Okay, that is now a scale challenge. So how do we make that happen? How do we enable maps, location and the user interface to work? Certainly with 1000 new locations. And so scalability, we focus on every single part of the problem. You asked about the maps part of it. With maps. What we've done is we've we've basically built our own self serve mapping solution and the whole intent of it Is that a retailer or any other organization can bring in somebody who is not familiar at all with our software, they've never used it before. They've never built a map before of course, and but they are very familiar with their local store, like they are the manager of their store, or they are, you know, one of the staff members at their store. And so we're making a Map Builder. So it can be super simple for them in a very short space of time to do all of the necessary steps to get their store up and running. And so, you know, we spent a long time in the very first store honing the user experience, we did that all manually ourselves, the next door that we've done, again, we did that entire process in two days of mapping and setting up the entire store. And that was a, that was a much larger store, you know, the next store that we will do will be a bit more remote. So we'll say to the staff members, hey, look, you go and you do this part of the process. And we'll do some of it in the background still, but you're going to do most of it. And we're not even going to visit the store. So you're going to be you know, in control of it. The store after that will again be more and more self serve. And so our goal really is to ramp that up and make that software super easy to use. So that anybody, any retailer can just say to all of their store managers Hi, look, you're going to go and spend a few hours or however long it takes to go and map this location. And it's really important that you do that. Because it's the only way to achieve scalability, it's a really hard problem to solve. And we've got just a genius mapping expert called James who's built this incredible mapping tool, which just basically takes a lot of the hard challenging problems, and just automates them away, or makes them super simple so that anybody can do it. So yeah, so that's one, that's one of the things that we've we've gone all in and we've done the investment on.

So you're offering a very holistic, complete solution to this problem. So rather than a one off a point solution, and really focused on selling your time, which is, as you described some of the other approaches that people are taking to getting some value out of trying to solve this problem,

I think of it is actually being a bit similar to stripe, right? Where stripe realized that lots of people wanted to sell stuff online, it was super difficult and complicated. And for all these financial reasons. And they just came along and said, Look, we've dealt with all of the technical complex problems. And we're just gonna like give you the simple interface. It's a thin layer of technology, which just enables commerce. And lots of people can now sell stuff online through that. And so we basically want to be the same thing for this kind of spatial technology. We can be this one platform, people can sign up, they can get it going really easily. And we can just take care of all of the specialists stuff, and they can build on top of it with everything that they're experts. And that's what we really want to be doing

to create this for any indoor space. Yep. And really easily as well, including my house,

not so much. No. So by focusing really on kind of law, I guess I define it as kind of public spaces, right? So think of that as shopping centers, airports, hospitals, museums, every type of retail store, I'm just listing off places that we get inbound from now casinos, and large entertainment, venues, hotels, that kind of thing.

Got it. That will be phenomenal. And you're as part of this vision, you today are building the user interface, the entire experience for your current grocery retail customers. Yes. Do you imagine that you're going to offer those as as tools for others to go in and take and adopt and build their own experiences on top? Are you going to be developing the entire solution for each one of these sorts of physical locations?

Yeah, you know, we sat down and talked about this recently as a team. Where do we want to sit? Right, like within grocery retail in particular? Why do we want to sit? What do we want to be giving to the retailer? Where do we want to draw that line to say, this is the bit that we're going to do? And then this is the bit that you're going to do? And really this is all defined by the market, right? Like, what is it that the retailer's want? And we have to discern that because they've never done it before. So there's a few bits where we're like, well, they don't know that they want this yet, but they're gonna want it. For example, one of the first things that we realized when we tested it was we did a to b navigation, right? You type in milk, and it just gives you the route to the milk. And then very quickly, we were like, Okay, great. Now the technology is up and running, we can be navigated to the milk. And now suddenly, you're like, oh, but what about the rest of the items that I want to get on my list? Right? Do I want to be manually typing in each item? And keeping in mind the order that those items? No, the app should solve that problem for me, it should do like multi item navigation. It should do it really easily. Without you ever thinking about it. Okay, is that easy to do? No. We discovered it's a really hard problem. Okay, great. Let's solve that part of the navigation piece as well, because that's going to be a key requirement for these for these retailers. Not many people are asking for it because they haven't reached that point yet. They don't have anything which we're But we just realized, okay, they're gonna want that. But there's many other things that we don't need to be doing right? We don't need to be implementing, you know, scan and shop solutions, they've already got people who are experts at that. We don't need to be building on recommendation system, they've already got people who are experts at that. And so what we've been looking at is, how do we build an SDK that we can hand off to a retailer, they can embed within the application they already have. And then they can just get going from that point on. Now, what we've done initially is, of course, we've built an entire application so that we can build some of those features and test them ourselves. And we can build a shopping list and we can build a nice search feature on everything, so that we can put it in the hands of customers and, and understand how people find it. But ultimately, what we would provide to a retailer is a bit more stripped back from that. And it's a bit about, it's more about enabling them to provide all of the value and all of the, the interface and their their own branding. And we just kind of sit as that kind of mapping navigation and AR solution on top of that.

Got it. So here we're gonna have our local grocery store app will still be the primary point of interface for engaging with that particular store. But now they've got all this additional list building and ordered item navigation through the store and all the AR elements are part of that app experience now,

right, exactly. The thing is, is that those apps are really popular, especially in the UK, and it's moving over to the US a bit more now. But just as one example, so Sainsbury's is the second biggest retailer in the UK. And then reported recently, it was even a year ago. So I imagined the numbers might be bigger now. But they have reported publicly that in the stores that they have, they're kind of scanning or technology 35% of their shoppers are using that app in order to check out in order to, you know, scan the items through the store journey, which is huge, right? So we look at the kind of distribution side of things. And we're like, okay, great, this is problem solved. Because if the next time 35% of those shoppers walk into that store, imagine if they could, or whichever retailer it is right and whatever their numbers are, imagine those shoppers walk into that store. And suddenly they've got this, you know, they've got the map of the store, and they've got this full navigation shopping list experience. That could be that could be really interesting. And that's something that we can just kind of read the existing success of

Yeah, wow, that is amazing. Maybe I'm not as avid of a grocery shopper, perhaps as I should be. But the scanning goes sort of technology sounds really interesting. And it's phenomenal. You get to ride on the coattails of all of the marketing and all the adoption of that bit of mobile app technology as well. Yeah.

And the really complimentary as well, because what we do, of course, is improving the shopping experience. That could literally be our tagline, right? If we, if we didn't want to tell anyone about like, what our long term vision is, we would just say is improving the shopping experience, right? And what those kinds of scanning things do. Or if you even look at Amazon Go and what they're doing. And there's a few retailers who are doing the same sort of thing as, as them. Amazon Go, of course, is the thing where you can walk in the store and walk out without lining up at the checkout, you just kind of put your items in your pocket or whatever you want to do. They've got a few of those stores in London, I surely tried to trick them up. And, you know, try and fool them and see if I can steal a banana or anything without them noticing, not steal it. But you know if I can get it off the shelf without them noticing it. But those solutions are all about the checkout part of the experience, right? They're about solving that bit when you have to line up and scan your items. And so I look at those as like those are super complementary, right, bring those together. And you've got something which is filling the store with information and guidance, making sure everyone can find everything they're looking for. And then they can check out really easily at the end of that. That seems like an amazing grocery experience.

As you think about the building out this kind of the team necessary to capitalize on the opportunity and the initial groundwork that you've laid. What lessons do you take from the earlier sort of game PokemonGo esque sort of experience, you're trading that early startup or? Or from all your engagement with the broader development community? What sort of lessons do you take from that in terms of building up the business and the team here,

it's really important to have a North Star that everybody's aiming at, and everybody deeply understands. And they think about every day, in terms of the development work that they're doing. And so I think it's really important to build a team that is focused around, you know, if the current goal is to build something really compelling and really valuable for shoppers, then every decision that people can take in terms of what they're going to build this week, or how they're going to build a particular feature or how they could improve it. They're always thinking about that. And I think it's important to have really fast iteration cycles and not overly build something. For example, we've done testing before where we've implemented things like recommendations and stuff like that where it's, it's as hard coded as you could imagine. And even one we've kind of done the hands on testing of that. We've maybe made people think that it's a lot more, you know, oh, look, something magically popped up on the screen. And we definitely didn't trigger it in some very specific way to make that happen. Because really, it's just about understanding, is that a valuable thing? Right? How did the person react to it? Did they get value from it? And if they do, great, let's go and build it properly. And so it's about having that kind of first iteration cycle where you can very quickly test ideas, see if there's value there, and then iterate on it, if it works, and then you know, do something else if it doesn't. And I think when you're building something, so new is what we're building, it's really, really important to do that, because it's super hard in a lot of different ways. CF. So having a very product focused team and a team who's just, you know, very ambitious and thinking about, how can we build a completely new type of technology that people are gonna love using,

is you think about building this team, this notion of having a North Star and the tremendous amount of value get out of these tight iteration loops, this tight iteration cycles? Yeah, how do you articulate kind of the core values that you try to imbue in the company or within the team?

We're a small enough team that, you know, we've all come from our own experiences of working for different companies, and we've all had good work experiences, or maybe some bad ones. And I think like, we've had this opportunity where we kind of go, okay, great, like, now we have our own company, right, we get to decide what the rules are, how we get to decide how we want to work together, and how we want to be treated, whatever the word is, in the workplace. And, you know, we may have had frustrations at previous places. So we just get to kind of define the rules for that and figure out how we would want that to be, we've got some really big ambitions, like working hard. And all of those things is like right at the core of what we do, we're super focused on a mission. And in order to achieve that, there's kind of two things that I think are really important two values that we've imposed, which again, come from some of the lesser good work experiences that I've personally had. And one of them is about collaboration. And this is kind of the sense that I think everyone on the team has a way more to contribute than whatever their specific role or title says, right. So if you're like an iOS developer, you might have some ideas about design. And you might want to, you know, say, hey, look, I've had some ideas on the design, I think maybe we should go in this direction, whatever it is, or maybe this could be improved in this way. And you should be able to input into into that. Or you might say, You know what, I'm a commercial person. But I've got some great ideas on marketing, I've been able to, you know, add some value previously. So like, I'd like to, you know, be involved with the marketing efforts a little bit, or trying some things out grid, like go into that. And so it's about fostering this environment of collaboration, which I think is really important. If you're hiring great people, they're going to be bringing multiple things to the table. And you want to be harnessing that as much as you can, if you want to build something really great. The second thing that is really important is about ownership. Everyone here has ownership over the area that they work in. And so the the key thing of that is, everyone knows that they have the trust, to go and make big decisions, to not just be in the box and told this is what you're going to do, and just unquestioningly do it. But just like be curious about it and be thinking, actually, you know what, we could do this in a better way. Or maybe we should go in this direction, because I think this will be better. Or maybe we should, you know, think differently about this thing, whatever the thing is, we want to encourage everyone to have ownership of what they do. Because again, you bring those two values together ownership and collaboration. And suddenly you have a team who will not just you know, it's not just filtered down from the top, everybody should do it, everybody is inputting everyone's being creative. And you're ultimately ending up with something, you know, be so frequently build features, that just confirm one person's idea that they had, that if we didn't have those values they might not have ever spoken out about. But those are some of the best features that we have. So I think those those kinds of values are really important. And of course, you know, we're a team of about nine people now. And as we grow further, we'll continue to iterate and develop whatever those work styles are, and figure out other ways that we can improve the way that we work to, you know, make us most likely to achieve our vision. And also do it in a way where everyone is like super excited to work here.

Early on, you were able to share a lot of the concepts that you were playing with on Twitter, and you got a lot of great feedback from that. And it really in some ways led to dent reality as it is today in terms of the sort of feedback. Are you still able to or how are you leveraging Twitter today?

Yeah, that's a good question. You have to be a bit more careful when you're working with big retail organizations. I've discovered. You can't throw out any surprises, right? You can't be like, Ah, look, we've got this amazing new feature. Because then the retailer is gonna turn around and go, What is this new feature like you haven't told us about this. So suddenly, you need to be a bit more careful about that. And that's something that's something that I've seen certainly learned and figured out is you want to make sure that whatever you're kind of seeing externally, it doesn't come back. And then the retailer, or even your stakeholder at the retailer hears from somebody higher up, he was saying to them, what on earth is this thing that I've just seen online, so you need to be a bit careful about that. But you know what we're working with a UK retailer called Marks and Spencer, I've mentioned, I've probably mentioned them a bit earlier in the conversation. But we've been incredibly lucky here where we set out straightaway. And we said, we want to start within grocery retail. And at the time that we kind of made that decision. This was in the depths of COVID, in 2020. And so we can travel anywhere in the world, right, we had to work with a UK retailer, of which there were about nine UK retailers. And one of them needed to say yes to us. Because at this point, everyone's focused on COVID. They're not like reaching out. And they're not thinking about new projects and stuff. So we needed to find a retailer who would work with us, who would allow us to go into their stores and test the stuff. And what we found with Marks and Spencer is just just this incredible openness to allow us to go in, you know, mortar stores, test directly with our customers have a lot of freedom of testing new features, and all of this stuff, and really just build this interface and ticket in an interaction that we wanted to take it in. And they've just been incredible to work with in that regard. And so even like as it comes to posting content online, they've also been incredibly open to just allow us to post stuff and test it and test in public basically. And so we've been really lucky with that. If we look forward to like how we work with other retailers, I think it's a really interesting question of how do we continue to show the internal innovations that we have, which I think people online will be super interested to see some of the amazing integrations and you can imagine all of the crazy stuff that we could do in other retail environments beyond grocery as well. Maybe some stuff that I already know about. And I can't talk about publicly here, like how do we go about showing that in a public setting? Without a retailer going? You know, what, you haven't even we haven't agreed to anything yet. You're just like doing a pilot with us. We haven't, you know, publicly announced anything. So yeah, there's so there's a bit of like a managing of that, where we have to figure out how do we go from being me as one person, like just posting videos of whatever I want to pose to actually no, we have a bit of a responsibility to make sure I clearly communicate with our customers. So it's, you know, it's a bit of a leveling up and making sure we're being a bit responsible about it.

You just talked about COVID, and how it adds an extra layer of complexity to starting a company at that time, especially creating a toolset that requires you to be in person in front of other people. Yep. As you as you look forward over the next 12 to 18 months, maybe COVID is now more in our rearview mirror, we hope, yes, who who are what concerns you most over the next 1218 months?

I think my biggest concern is the heart of everything that we do is scalability. And we know that we've got this backlog this ever growing mountain of it's a really good problem to have. In most startups. I know, most startups don't have this. And we're very lucky to be in a situation where we get an incredible amount of inbound from all of these people who would love to work with us. Now the challenge there is just about getting ourselves up to the point of scalability, where we can open it up and we can say, Yeah, sure, go ahead. Like we get an email from someone tomorrow. And they say, Look, we're kind of interested in this. This looks interesting. Could we play around with the SDK? We can just send the mango? Yeah, great. Here it is, here's your access. And they could go and do it. Right, we need to be at that that point. And right now, because there's a bit of hands on requirements from our team in order to map every new store, work with a new commercial customer, there's a few steps remaining before we can really open it up. And so for me, it's about every day that we're not making big progress. And scalability is another day that we're not able to capture some of this opportunity that's coming into us. But it's kind of being blocked behind this door of non scalability, basically. And so you know, so I think if I'm being fair to ourselves, we are way ahead of everybody else is a comes to scalability. But we are still not at the level we need to be in order to unlock all of these new customers. So there's, there's a bit of work we need to do this new people we need to bring onto the team like integration people and, you know, account managers and stuff like that in order to enable some of the stuff that happened as well.

For you, yourself. You're, you know, you came at this problem as an expert developer. Innovating Yeah, in creating some solutions to some very hard technical problems. And I'm curious about a couple a couple of things. One is how is it that you, you yourself have evolved or changed the most as a leader Over the last couple of years being the CEO of the startup,

in every way possible. Before this, I was a developer, I was a lead developer. So I'd love to see a small team of people had done that a few different jobs. But when you become a startup founder, you have to learn about all sorts of new things. Basically, if you're an early stage founder, half of your job is like, here's some, like, every day you arrive at work, and it's like, or metaphorically arrived, because I'm not I'm not traveling anywhere, but you open up your computer and like, half of your day is spent with like, here's a problem that you've never seen before, go and figure this out, like, figure out how to solve this problem. And you just have to do it. And you have to learn about whatever the area is. So I mean, I've learned a huge amount about raising investment, interacting with investors, about finance, about legal stuff, about trademarks, patents, about marketing, about PR, and working with people like TechCrunch, to get yourself featured and how that all works, about team management and trying to figure out how you build a team, how you manage, like, how you manage the finances, right? How you manage a roadmap, like you say, great, well, we've got, you know, until this date to achieve this, and here's what our roadmap is going to be. And here are the team members we're going to need for it. And so great, we need to figure out how to hire all of these people have never done anything like that before. So there's just all of the stuff that you just need to become an expert in basically, not an expert, but good enough that you can get away with it. So yeah, I mean, if if I was to kind of look back at like, myself, as a developer, before this whole thing began, there's just a huge amount of difference and a huge amount of stuff that I have that I've learned over time. But yeah, it's a fun experience,

is there been an entrepreneurial side of it has it been anything that is substantially harder or different than you expected, it would be

probably raising investment, depending on how the round is going, it can be raising investment can be, it can feel awesome. Or it can lead you to a place where you're like, This is literally the number one risk to my company, is that these is just these working with these VCs or whatever it is, right? And so like, it's, it's something that you don't have a lot of control over, like when you have to raise your after raise. And it can either go really well or can go really badly. I've certainly had both sides of that situation. We were raising we we our plan was as we roadmap to out in 2019, right? We were like, okay, brilliant, we've got this new funding, it's we we just raised our precede from angels, that was an incredible raise. And then we said, great, so this is going to take us until our next raise, which is going to be winners that are March 2020. Brilliant will raise money then. And then we went into that, and that was obviously just awful, because suddenly all of the doors were closed, you know, you couldn't raise investment at that point. And so that was that was a really bad experience and really scary. But we you know, we basically relied on angels, and got through that, that period. And now we've had really awesome, amazing experiences where just you build momentum, and it just builds and builds, and everybody wants to invest in you have to turn people away. And that's when when it feels really great. But overall, I'd say like, if I was able to just magic wand, get rid of one aspect of what we do, I would say probably investment, because even though even if it is going well, you still think, Okay, well, next week, either I can work on, I can speak to investors and sell them on what we're doing and convince them it's a good idea and have the opportunity and that they should be involved. Or I can have the second or third call with this person. Or I can work on this new feature that we want to add to the product, which we think will increase the likelihood of success or whatever. And so you have to make real decisions on we're not going to build this feature, because my time as a technical founder is going to be spent on raising money and speaking to investors. And that's always like a painful thing, even if it's going really well to have to kind of make those calls. So yeah, if I could cut one thing from the process, it'd be raising investment.

When you're not busy raising investment. Congratulations on closing the relatively recent round. Yep. Thank you, when you're not spending time raising investment, how much of your time do you get to spend on the code itself?

About 50%, which I think is pretty good. I spend some of my time doing, you know, stuff that I just don't think is a great use of my time like the I'm like, I can do this. But I can do legal stuff. I can work on patents, I can work on getting people visas, but it's not the best use of my time. The best use of my time is like on core technologies, on working on product working on special interfaces in new ideas. Like that's where I feel like, you know, I can maybe add some value there. But yeah, I get about 50% of my time on it, which I think is pretty good. And a good balance.

Do you have any special approaches you have to managing the stress? Today?

No, I'm really bad at it. I'm really bad at compartmentalizing, I should be much better at it. But it's I find it really difficult. Actually, when you're kind of like, you're like, Okay, great. I'm in development mode headphones on, I'm going to do this for the next four hours. And then just the back of your mind just enters maybe seven or eight different things, fires that are going on fires that, you know, you will put out when you get to them in the afternoon, and you start are the situation I'll move it forward and everything. But like, you just can't help but have it enter your mind. And that might be solutions to that I would love to know what they are. I am not particularly good right now, compartmentalizing. And that's one thing that I need to improve on. Because it kind of it's one thing that I really love about working on the weekend, which I which I try not to do every weekend. But when I do work on the weekend, it's great because it lifts that stress, nobody is expecting me to to complete any tasks or get back to them on anything in so I can really focus on development or ideating on something or some aspect of code. Because I know, okay, great. All that stuff is genuinely on pause, and I can get to it next weekend, come Monday, all of that will, you know start to ignite again, Ignite back into flames. Not quite that bad, hopefully be during the week, I find it quite difficult to focus to focus on stuff. So I could be better at that.

Have you tried any sort of meditation practice? No, not super seriously. Now, it's a technique that I found works well, for me, especially when I'm this feels like there's way too much going on is to actually take a step back and spend 10 to 15 minutes working on some sort of mindfulness meditation, which allows the next many hours to be a little bit more focused, I find that at least personally, I'm able to ignore the deluge of of interruptions and inputs and everything else. And sure, in which my mind to the one thing that I really want to focus on for the next couple of hours, it gets a little bit easier, a little bit easier with a little bit of a mindfulness practice,

right. You know, I always kind of try out things on a weekly basis where I go, where I go, you know, let me just try this for one week and just see how it goes. So maybe I could try that. I'm not so big on like New Year's resolutions, because I feel like those are very born to fail, right? It's like, oh, it's a, it's a big cultural thing. Everyone's talking about it, what is my resolution going to be and then it gets less popular, and everyone gives up a few weeks after that. But I but I kind of do those things on a more, you know, just like as they come during the hour, maybe I could improve in this area,

just like in software development, continuous iteration. Right. Exactly. Yeah. Let's wrap the view in lightning round questions. All right. What commonly held belief about AR VR spatial computing Do you disagree with

I think there's too much focus on marketing gimmicks, and and those kinds of ideas of people go weigh heavy on the idea that you'll get like, you'll hold your phone up to, I don't know, like a box of cereal, and it will come out with some flashy animation and all of those things. And maybe there will be a future where some of that exists in the world where advertising becomes more 3d and more dynamic, and that sort of way. But the sheer ratio of content, which is that to like, utility is insane. And so I think, yes, maybe that will be a feature part of things. But I think it'll be a lot more pared back. And I think right now, there's people just there's so many kinds of AR companies who kind of start with that, and they go great, we're going to be like this advertising platform for AR or for 3d. And it's like, great, like, maybe you want to work on the use cases first before you go deep in that area. And I think maybe it comes from just kind of thinking, Well, look, you know, Google ads is applied to all websites. And AR is basically about building that kind of interface you could think of like, what we're doing is basically building like the, you know, the digital homepage, that's a cringy analogy almost, but have like a location, right? It's like, this is the digital interaction someone has with some somewhere. So therefore, okay, you could put virtual AR advertising there. But anyway, I think he gets way too much focus. And it seems to be like something which I think is detrimental to the space as well, because then other people looking at all of that stuff. Think Oh, AI is just like a gimmick technology. And it kind of detriments the overall image.

Yeah. I don't imagine that most people look forward to a future in which they get spammed with more advertising right already are but now it's full resolution, very immersive 3d. Yeah, definitely. Really focus on the utility. Yeah. Besides the one you're building what tool or service do you wish existed in the AR market?

Well, I think actually, it's about the you know, what I talked about in terms of interface, because if you're an AR developer, and you want to display any type of information within the world, this does not have to be navigation doesn't need to be tied to public spaces. It could be, you know, an app which, let's say you're visiting an Airbnb, it integrates with the Airbnb, and it shows you like where the thermostat is and how to use it. And you know, where you can find the the kettle and other utilities or whatever, right. Or maybe it's a cooking app that displays on top of your hobby and shows you the timers. Or maybe it's one of those piano apps, which I think will be brilliant with wearables, by the way, where you can see the virtual notes coming down. Yeah, it'd be amazing. If you're a developer of one of those applications or any other type of application, there's going to be a need to display information tighter to those virtual objects. Or even if you're doing you know, that demo that I did early on, like the book demo, where it recognizes the cover and shows you information next to the item, you're going to need some sort of framework to be able to lay out that special interface to be able to show that information within within the AR space in a way that makes sense. And right now, when people do this, they find Okay, well let me just make a virtual label object. And I'll display that in the world. And then they realize, oh, actually, it's way too big or too small, or it should behave a bit differently. And so everyone independently is figuring out all of the paradigms of how that interface should work. It's like, imagine if Apple had released the the iPhone SDK. And they didn't have any support for scrolling lists, or even like pinch to zoom, or any of the other multi touch technology, or they hadn't even given a guide to how like how, how our navigation structure should work. So like, here's the back button, here's like, a button. Here's what a button looks like, right? So imagine if they hadn't provided that interface. But they just said, everyone can just independently go ahead and like build what you want. On this touchscreen, everyone would just go ahead and like, independently figure out how to build that. And some people will do better than other people. So what I think I really need to somebody to kind of come in and be like, and I think probably should be the platform provider, to be honest, who will come in and go great. This is like making it super simple to display regular content within the ER space in a way that makes sense and is cohesive, and works. So that you can just focus on building your application. And you don't have to worry about all of those really deep, technical aspects of doing that.

That will be amazing. Do you know who I remember being the great pioneer as it relates to mobile interfaces? It was the palm team building, I think it was Web OS. All right, back in the day, they were the true geniuses who had figured out so much of what we take for granted today in terms of mobile interface,

was it called like the palm tree or something like that, for device

at that time, there was a palm trail. But if I recall, the Web OS team, I think there were, maybe they were making it for the trail or whatever, whatever their version of the the iPhone like device was going to be. But I remember after this kind of amazing presentation, and amazing work, that team had done that Apple did a brilliant job of borrowing all of the best ideas.

Yeah, and it's incredible. Because I've seen that video, I didn't see it at the time, but I've seen it since. And even stuff that they had, for example, the idea that you could like, interact with it in such a way that you could flip up a card of every application and they could be open. And you could flip between them and then dismiss them. Much like the way that Apple brought that to the iPhone, with the iPhone 1010 years after the original iPhone. And they were still bothering those, those ideas in those interface concepts from the Web OS development team from over a decade earlier. So yeah, it's incredible. Like, just seeing like, how pioneering even if the device didn't take off? How pioneering that work was?

Yeah, similar to the work at the Xerox research group, right, this park. They designed this all of these concepts around windows and graphical user interfaces that were heavily borrowed by Microsoft and Apple and brought to the masses, similar with this web with

us has been portrayed in numerous movies. Yeah.

What book have you read recently that you found to be deeply insightful or profound?

Yeah, this is going to be a bad question. Because I recently in fact, I'm currently reading the book sapiens. And this is the most generic answer possible, like it's the, you know, probably go to thing but obviously, sapiens, for anyone who doesn't isn't familiar with it. It's just, it's, it's about, you know, the history of mankind basically. And I just think like, the first half of the book, or the first third of the book, I think, is the most interesting, because that's when it rapidly goes from the scale of kind of billions of years and then into the 70,000 years ago, and then it rapidly takes you through time of the evolution One of humankind and the ratio of really interesting, ha moments per page is really high. Like those moments, they just go, Huh, that's interesting. I didn't know that. Today I learned the ratio of of those per page is really high, and that the first third of that book, and then after that, it starts to go into deeper topics. And it tends to be a little bit more drawn out and in comes into the modern day, or it's a bit more like about the scientific revolution and stuff, which I find a bit less interesting than finding out what happened 10s of 1000s of years ago, and how things like how it came to be that we are the only how Homo sapiens is the only type of human right, why is that when you get like, numerous types of cats and dogs and things like that, how come we're the only category of human well, because there were numerous ones, and we killed them all. It's pretty much the consensus on it. And you're like, how did they know that? Oh, it's because they can look at the archeology in the bones. And they can see, well, look, the Neanderthals existed in this one location. You can see how old the bones are, and they go up to a certain period. And then they just stop existing. And then the the HomoSapiens bones start existing at the exact same time. And they've since seen that play out in hundreds of different locations around the world at different points in time. But yeah, there's a lot of things like that where you think oh, actually, yeah, I hadn't thought about that. But really interesting insights.

I really enjoyed that book as well, especially that part. I remember in that aspect. Two things, there was one extra element that really stood out for me, and that it wasn't just that we had killed off the other species of Homo whatever's it was that we incorporated some of them into our selves, right, there is some element of Neanderthal DNA. In modern Homo sapiens. There's some Denisovan, adenosine Aryan play mispronouncing or mainly particular variety. Also, within humanity, like we have these we've incorporated also some of the rights are there are some inter marital stuff going on. Or racial? Is it right word? interspecies. Yep, cohabitation that was happening. Anyhow, if you could sit down and have coffee with your 25 year old self, what advice would you share with 25 year old Andrew?

I think if you're like an independent developer, an indie developer, which I'm not so much anymore, but I've got a startup, I've got funding and stuff. But for a while I was kind of building stuff on the side. And what you really need those, you need to be surrounded by people who are like, have a similar mindset to you. Because I have a lot of drive, I have a lot of energy and motivation and stuff. But I think it would have like been helped even more to kind of like, be surrounded by like minded people who are also doing the same sort of thing. And so what I found over the past few years is I've just found those communities, and I found those groups of people who were really like, able to, like motivate me and be interested in what I'm doing, and really get to the idea of being a kind of like, indie developer, and like, Oh, look at this awesome thing I built over the weekend. And then you can look at their projects and be like, Oh, that's awesome. And you can really just motivate each other. And I think, you know, because I'm kind of seeing this as though like, imagine somebody is, you know, in my position that as I was when I was 25. And like, what advice could I imbue there is like, try and find those communities of people and try and surround yourself with those people where, when you tell them about what you're doing, they're just like feeding back that positivity. In it's, it's great. I think they don't even need to be geo look local to you anymore. Because there's all sorts of discord communities that you can just join in the filled with all sorts of people with the same sort of mindset. And so that's something that I found super valuable. And over the past few years, but I didn't like that wasn't a thing in my life, when I was actively building my own projects about five years ago, or a bit longer ago. And I, you know, I think it would have been beneficial. I think the other thing is, is that I would spend a lot of time building out applications, which were absolutely beautiful, and add every feature that I wanted them to have, and all of these things without doing really rigorous user user testing. And thing is, is that I'll give you like a free app idea, because it's an app idea that I had recently. And I was thinking about how would I do this differently today if I was building a new app, versus how I would have done it five or six years ago. And the idea for the app is basically whenever I watch a TV show, an episode of a TV show, I always kind of feel like oh my god, that was an amazing episode. I need to like, discuss that with somebody like I need to like, you know, find a community of people to talk to this about and there have been a bunch of apps over time which have tried to do this and basically knock it off. To the ground, because it's really hard to build a community like that episode or content just happens once and then it's gone. So how do you build that community over time, especially if you're watching something that was released many years ago. My solution to this is after I watch one of these episodes, I'm watching the TV show Silicon Valley right now. And so I might watch an episode and I might go, that was an awesome episode. And then I Googled, like Silicon Valley series, season one, episode eight, Episode discussion, Reddit. And there was a Reddit thread for every episode, have basically every TV show that is on right now. But you have to like go on and manually find them and just use the Reddit web browser to read through them. And so I think what you could do is you could bring that you could pull that into an app, right? So an app which basically organizes all of these TV shows and episodes. And so you can very easily filter in and just quickly find the episode you're looking for, and read the discussion. And you don't have to like manually find it yourself online. So that's an idea for an app that I've had, if I was making this years ago, what I would do is I would make that functionality. I would spend ages and ages building the absolute perfect system of moderation and figuring out how am I going to fill in the data with all of these TV shows, maybe I would build some AI algorithm, which is super advanced, maybe I would, I don't know, like have all sorts of like edge case stuff built in, I would have user accounts, so that I can track which episodes of TV shows people are watching all of this unnecessary stuff. If I was building it today, what I would do is version one is I would have like, a really simple thing, I'd be like, Look, this supports 10. TV shows, I've manually gone through and added all of these episodes. And I would just basically say to a handful of friends, what which TV shows, are you watching? What TV shows? Could you see yourself using this for? Okay, great, I'm gonna manually add in those episodes. And so here it is, like, use it, see what you think of it, there's going to be no user account, it's not going to track your progress, you're just gonna get like one feature out of it, which is being able to read these threads and access it in a nice way. And then I would get the feedback and see if they like it or not, maybe they would say, this was not useful to me at all, this is the worst app I've ever used. Okay, great, I'm gonna, you know, get rid of the idea and forget about it. Or they could go, this is really cool. I wasn't so sure on this feature. Or it will be you know, I had this problem. And then I can, you know, speak to a few people and see if they say the same thing. And then I can figure out how to build a feature that solves that problem. So basically, my approach to it will be a very much like, lean mentality of build the minimum thing. Whereas like, if you're an indie developer, the number one temptation is to ignore all of that and build the most beautiful feature filled up that you've you can ever imagine, spend a year and a half on it, and then release it on the App Store, and then discover all of that feedback. So yeah, that was a very long winded answer to your question. But yeah,

yeah, no, it makes perfect sense. The lesson is that I think if it is buying down risk, you want to make sure that you're not building too much in the wrong direction, right? Because chances are, you're wrong about something important. Yeah. So you're better off finding the minimum sort of thing that you can go and build and confirm that you're going in the right direction.

Yeah. And I think it's easy to do it if you're in a job, or if you're backed by investors or whatever. Because you're like, Okay, great. Like, I'm taking this as a serious business opportunity. There are external reasons to my own enjoyment, why I would build this in such a way. Whereas like, if you're an independent developer, the temptation is to go, well, this isn't a job. So I'm just gonna, like, this is my passion project. So I'm gonna go and build it and everything. And there's nothing wrong with that. It's just that ultimately, you're going to be so much happier, if this is an app that loads of people really like using. And so you really should just import a little bit of that, you know, proof of concept mentality and opening it up a bit earlier. And like, there was if the result of that is a year down the line loads, people really like what you're doing, and it's, it's great, then, you know, fantastic. That's going to be a much better place to be.

Yep. Any closing thoughts you'd like to share?

Yeah, there's one thing that I should definitely mention, which is about hiring. Hiring is a big thing that I'm focused on right now. We have an incredible team, we need to expand, we need to bring on new people. One of the main roles that I'm looking at right now is the the operations are all right, whether you call it a CEO or head of operations, we basically need someone to come in and be a day to day on the ground problem solver and help us to really build and meet all of the scale needs and figure out you know, unplug a bunch of problems and figure out basically how do we get to all of these, all of this inbound customers. So we're looking for the right person for that role. So if anyone is listening and you think that could be something I'm interested in, or you think you knows Somebody who can be right for that role, then, you know, go on our website careers don't realty.com. And you can find the role and you can apply there or you can just message me on Twitter,

or do both. So where can they find you on Twitter?

So my Twitter is Andrew Hart, ar, heart is spelled H AR T. Maybe I should have just capitalized the A on my surname, and that would have been enough.

You can always create another account, maybe? Yeah. Does it matter where this person sits this new head of operations or see oh,

yeah, that is important. So when we are looking, we're building a team in London, so the person needs to be in or nearby

London. Got it. Great. Andrew, thank you very much for this conversation.

All right. Thanks so much.

Before you go, I want to tell you about the next episode, and I speak with Kai Stroder, co founder and CEO of tooz a company on a mission to develop cost effective smart glasses for consumers. Their injection molded plastic lenses can be curved and made into stylish glasses that incorporate vision correction. Kai shares the story of how tos was spun out of the Carl Zeiss smart optics group and how they're creating an offering you'll be able to buy through your optometrist. I think you really enjoyed the conversation. Please follow or subscribe to the podcast. Don't miss this or other great episodes. Until next time