The Stack Overflow Podcast

Crafting software and games for the selfie generation

Episode Summary

This week we chat with Tara Reddy and Sam Weekes, the co-founders of Loveshark, a mobile gaming company. They describe the company as a team of creative, kind, and slightly quirky people who are building the future of social games. They use technologies like augmented reality and computer vision to create weird and wonderful things.

Episode Notes

You can find Tara on  Twitter here. 

Sam is on Twitter here.

You can learn more about Loveshark's latest games and the roles they are hiring for here.

Thanks to our lifeboat badge winner of the week, Elliott Frisch, for answering the question: Convert list of integer into comma separated string?



Episode Transcription

Tara Reddy Yeah so we've really focused on how can we bring play and sort of social together? I think there's this just amazing spaces sits between social networks and games that's not been fully explored yet. You know, we see the starts of it with you know places like Roblox and and Fortnite, but what about something that's mobile first that utilizes that selfie camera that everyone's using every single day. And that's maybe built in a way that's catering for a different audience.

[intro music]

Ben Popper Build your health app on the most accurate data available. The Healthcare Locator SDK instantly connects your apps to the world's leading health care database. Add provider names, locations and specialties in just hours. Visit to download today. 

BP Hello, everybody. Welcome to the Stack Overflow Podcast, a place to talk about all things software and technology. I am Ben Popper, Director of Content here at Stack Overflow. And I am joined this week by Tara and Sam from Loveshark. Loveshark is a company focused on making cameras games, and I read a little bit about that one layer deeper. And it seems like it's kind of playing in the world of AR augmented reality and also social. So Tara and Sam, welcome to the show. 

TR Hey, Ben.

Sam Weekes Thanks for having us. 

BP Yeah, my pleasure. So I guess in your own words, we'll start with you Tara, tell us who you are, yeah, and what it is you do at Loveshark?

TR So I'm Tara Reddy. I'm the CEO and one of the co-founders. And I cover everything to do with product. And but I also do marketing, ops, finance, just making sure the company runs as well. 

BP Okay, sounds busy. Sam, how about you?

SW Hey, I'm Sam Weekes, I'm the CTO and the other co-founder of Loveshark. I kind of cover everything from the tech side to design, product, everything else in between.

BP Okay, very cool. So yeah, I mean, this podcast is typically focused sort of on the art and practice of software and tell us a little bit, I guess, Tara about the inspiration for it. And then maybe we can dive into kind of like the stack, like, when you started building it, what were you using? And these days, you know, what does it take to create a great game or a great experience?

TR Sure. And so Sam and I met when we were working at Blippar. So Blippar was one of the sort of like AR unicorns of, I guess, the 2010s. And they had a mighty, mighty rise and fall, but they were they were huge. And they covered everything from computer vision to augmented reality, they built some of the first AI engines, you know, they did b2b products, B2C products, they just had such a wide span of everything to do they are. Sam was an engineer there, I was a product manager. And we work together on several social products for Gen Z that incorporated AR. And we just thought there was this massive opportunity to apply these technologies to games, and we just didn't see much out there in the market. And I guess Sam and I realized that we actually had such a deep experience in like creating and launching these type of products. And we felt like we would have an edge if we went after that opportunity. So after you know, a few years at Blippar, we left to start Loveshark, we built our first game in five weeks, launched it, and it got featured by Apple straightaway, which was amazing, just validation that you know, there was something there in the camera game space. And then from there, we've went on to raise multiple rounds of funding, grow the team. And now we're building out our first sort of camera game platform.

BP Cool. So actually, it's interesting. We didn't set it up this way. But we happen to have the folks from Niantic Labs on like a month or two ago. And you know, similarly, they were talking about, yeah, the wide sort of universe of applications, everything from, you know, I want to see how this lipstick might look on me with this color. I want to see if this couch would fit into my room. And then obviously, you know, they have a whole set of games and then you know, a platform to sort of support it. So you know, I want to hear more about Loveshark but I'm just a little curious, since you said it, what the rise and fall, like what was Blippar doing that was working and what caused it to not work or like was there something that they were focused on AR that didn't end up being the main use case, or they just get kind of ahead of themselves? What happened there?

TR I think the main thing is probably timing, to be honest. I mean, I feel like AR is only just starting to take off in the last you know, what, like, two, three years, and Blippar what started nearly a decade ago. So they were very early to the market. They were first movers, they were creating all the technology. And they were they are far before like Apple and Google releasing stuff into the ecosystem. So they probably served as a good example and a good sort of learning place for other companies to sort of like copy and refined to be honest.

BP Alright, so tell me a little bit about your first game. You said you built it in five weeks? Yeah. What did you use to build it? And what does it take I guess, to test something like that out and then you know, when it gets out into the real world with physics and occlusion and a bunch of different mobile devices, you know what makes it work on the consumer end but let's start with what did you you know, what, what did you use to build it in that five week run?

SW So the first game we built was called Laser Draw. So it's essentially like Pictionary. But in AR. Rather than just basically the two of us could basically sit down with each other, both of us in the same AR space, I could draw, and Tara would be able to see exactly what I was drawing in real time. All of that was using Apple's initial version of AR kit, we spent a lot of time actually building the tech around being able to communicate live between both the devices, whilst running the same similar AR experience. It was a lot of guesswork and a lot of trying out and testing and playtesting between Tara and myself, literally in the offices we sat in, but then also bringing in others to then come in and like just try out and see how they felt about it. Because as we were kind of in the weeds and involved in that project, you can't see the bigger picture, sometimes of like oh, this issue being like kind of being raised while you're looking at it too much.

BP And so if you're using AR kit, what languages and sort of frameworks you working with? Is it all stuff Apple has built? Is that like a higher level of abstraction, you're sort of playing around in an editor that they provide to you?

SW So the vast majority of the work we did was, we did most of it programmatically, so it was all within the Swift ecosystem. AR kit has their own, like they've got their own add ons to the normal Swift UI that you can see in Apps generally. But then they have a scene kit renderer, they have their own life, they've now actually started bringing out AR versions of that scene cam renderer is very similar to something like Unity, but it's because Apple at that stage weren't really focused on like games or anything like that their rendering engines for testing and building were quite low at that point. And then it started to increase and get better as time has gone on.

BP And so you said you've done a few different projects since then. Is it all in the game space that you released after that first one?

TR Yeah, so we've really focused on how can we bring play and sort of social together? I think there's this just amazing spaces sits between social networks and games, that's not been fully explored yet. And we you know, we see the starts of it with you know, places like Roblox and and Fortnite, but what about what about something that's mobile first, that utilizes that that selfie camera that everyone's using every single day, and that's maybe built in a way that's catering for a different audience. So just dropping something in there, our audience that we build for is Gen Z females. So basically, teenage and early 20 girls.

BP So if I go to Snapchat, and you know, I'm using these AR things, and I'm creating messages and sending back and forth and being playful, maybe I'm starting to key into that world. And some of that, like, activity that's already happening organically. I feel like once or twice in the past, I've seen something that kind of like, dabbles in the idea of a game, you know, like, you open your mouth and the snow cone comes through or the pie and you have to catch it. So what are you working on? Is it in the same sort of mind in that same vein of activity that's happening that like I would see on a Snapchat?

TR Yeah, I think we build on what Snapchat has already taught users. So we've been lucky that Snapchat and TikTok have started to teach people about how to interact with AR, because actually, it's such a new type of user experience, that it's actually quite taxing to train a new user and for them to, you know, pick it up really quickly. So luckily, they've started to do the training for us, the sort of stuff we build couldn't necessarily be just like building Snapchat right now, because they don't allow, like such a deep level of interaction. But yeah, I think it builds upon what's been started there.

BP And you mentioned when you started out with this first game and AR kit, you were working with Swift and sort of the tools Apple had since then, have you gone to Android or to you know, something that's more cross platform and have you used the same set of technologies?

SW Our second game, we actually moved to Unity. One of the again, because AR kit was in its infancy when we first started on our laser draw game, being able to do like, the usual kind of workflow that game companies do, where they, it's not just developers using that particular tool, we have like the designers involved and being able to like update and change things on the fly in their actual, like Unity editor type thing, we actually moved over to Unity for that reason. We tried it out, it worked a lot better. We were able to non technical people were able to actually update and improve on what we were doing help do the design work and everything. But ultimately, we've then come back to our latest projects, our camera platform, and over the course of iterating between not just from our like tech stack, but also from the product itself in terms of sizing and all of that stuff. We actually decided to move back away from Unity and back into scenekit and spritetkit and actually just stay within the native layer of Apple's framework. The main reason we didn't go to Android is because I'm not an Android developer, and I was one that was built in the beginning. The other reason is that we've always found with especially like someone at Blippar, Blippar was on both platforms. And we always saw that being able to try and create new features or new, completely new versions of apps, you have to count for both of them. And it means that progress doesn't happen quick enough. Whereas on a single platform, you're able to go a lot further a lot quicker.

BP And Unity is C sharp. So that's, that's like what you're using when you're working on it?

SW Yeah. 

BP So you mentioned at the beginning, Tara, that you were also thinking now right about sort of creating a platform yourself and then letting others build on it. So what are your thoughts there? Like what are the kind of tools you would offer to folks who want to work in the AR space that you yourself, you know, would have wanted when you started or still want? Like, what can you offer at the platform level that people can't get, yeah, just, you know, working with it sounds like some some pretty well established, you know, the Unity's and the Apples and the AR kits of the Google's folks like those, you know, that they obviously have big platform scale. So what are you going to build that folks can't find elsewhere?

TR So we really want to open up those creation tools to anyone. So it's not just developers or designers who can create, you know, games using ml and an AR. So we actually are building something where any kid in the world could upload a video, and it gets automatically made into a game, they might have to like, do a few little tweaks, and do some settings. But really what we're trying to open this up. So it's as easy as making a TikTok to make a game.

SW We want the games to be as easy as sitting in a bedroom and coming up with a dance while listening to music. Like that's as easy as we want it to be.

BP Okay, I see what you're getting at here. And I have two young kids, they're going, they just finished, they're in summer vacation. But they'll be going into second and third grade. And you know, they're obsessed with Roblox and they play, you know, games in there every day that maybe another, you know, second or third grader has created. But it seems like yeah, the process for creating games, there is a bit more involved, you know, like you can, it's easy to enough to play the game. But when you want to actually do it, there's a bunch of Doc's, there's a bunch of tutorials, you've got to get something on your machine. And obviously, like we talked about earlier, you know, people are already sort of like, yeah, for every teenager experimenting with the form of what's possible in AR and like, what's going to get me the most engagement is a natural activity. So how do you go from, yeah, like you said, you know, I'm just using the camera to that creates a game like what are the rules? And the points come in? You know, where does this, how do they add structure, I guess to you know, the activity of using their camera the way they're already using?

TR Yeah, I think I mean, games are really difficult to make even look at like game companies and game studios like there's so many that makes games that are just actually not fun. Or you know, they sound they sound so fun. But when you actually put them into practice, you know what I mean, it doesn't always turn out being as fun as you thought it would. So, you know, when you extend that to just general public, yeah, it's going to be even harder for them to make effective games. But that's where I think we need to do all the legwork. So we actually need to find a structure that they can sort of like work within. And I think the easiest way to do that, at this point is for us to perfect it with a game internally. So you know, we're making a dance game at the woman, once you know, we get players on and they test it out. And we can see that it's really working, we flip it and say, okay, you can make your own levels. But by playing, playing the game, you know themselves so many times, they're already intrinsically trained on what the rules are. So it's sort of like, I guess you turn your players into creators, and they're already trained by through the art of playing. 

BP Yeah, I mean, that's kind of how people learn, right? Like, if they're going to build a mod or a new level for a game, they've played it so many times. And they just think like, oh, man, I'd love to build a level that looks like this. So yours is like, can we talk a little about the game? So I have to do a certain set of dances and touch a certain set of spaces. And then once I realized how that works, then I can go into the space and do my own. And then people have to copy me, how's it gonna work? Are we allowed to talk about it? I know it's not released yet. 

TR Well, we'll tell you what it is right now. It's a product called Mochi. It's a dance game for mobile. And it uses machine learning to track your body movements as you play, and score your dance. And then at the end, once you've finished your dance, we add AI track body effects to make a video that's then shareable. So there's that sort of social networking element added on top. And essentially, what we really need to do is, Mochi is in soft lunch at the moment, it's in a couple of Asian geographies, and it's coming to the UK very soon. And once we validated it here, we will be, you know, going to the US as well. And it's just a case of like building a really strong community around that first, building the technology for the game maker, and then introducing that to the audience.

BP This is interesting. So I guess like you mentioned, there's a little bit of machine learning in there. Is that something that you know, you have to have, you know, at your company a competency in and you had to do training, is the stuff you're pulling off the shelf, like we talked a little bit about AR and what kind of technologies you use, but when it comes to doing software development that involves machine learning, how do you add that competency to your company?

SW So we've actually got our own machine learning engineer who is absolutely amazing, and just shows me all these random cool, amazing things that he's done. We've built on we have built on a few things that have already come out. So like things like pose estimation, gesture recognition, a lot of the times the techniques are already out there. It's then just training models to be able to be used by what we need them for. The big thing that we've always tried to do is, again, we've always been that user focused products. So it's always been, if the user experience is good, the tech can be a little less powerful and a little less accurate. And so what we've done is we've worked hard to be able to get all of our technology running in real time on device to be able to be not just where a lot of the traditional like computer vision techniques, kind of obey the post process action. So record the video, put it into the model, and then it will spit out something at the end, we've actually been able to get it all running on device at the same time, like real time, sorry, whilst the person is playing the game, which means that not just like, they can get instant feedback from their gameplay like, oh, yeah, like, you're actually doing a lot better, you're hitting the right moves at the right times. But also, it allows us to then overlay the AR experiences, because that's been powered by the machine learning stuff. And it means that they can then enhance their videos, because they're able to see what's going on, you get that like, Oh, this is amazing. And it makes them more expressive.

BP Right. You know, when you look at the world of AR, you know, having had that experience at a very early, you know, pioneer in the space who maybe was a little too early. Now doing it yourself and seeing other folks, you know, bringing on tools wanting to bring on you know, sort of the tools yourself, you know, you're building the picks and the shovels and people gotten used them. So, you know, you get to that stage of it. What do you think is gonna happen in the next couple of years? Are there things that would make a huge difference? Like if we got, you know, glasses that suddenly actually, you know, deliver that experience? AR would take off like a rocket? Or, you know, are there technological limitations that are holding us back in terms of, you know, the physics in the space, or the ability to do mapping? Or yeah, I guess a good a good one here would be like tracing somebody's body being able to like capture them and put them into AR in real time for a game like the one you're talking about. When you think about AR and where you are right now, what are you excited about? What's holding us back? And where are we going to be in five years?

TR So in my opinion, I think that it's it's a content issue, rather than a tech issue. For us, we've always sort of built a little bit ahead of what the techs able to do. And the tech always catches up. So if you can work out the UX and the core experience of what you want to do, even if it's not perfect, we find that the tech always catches up. And I think most people that are, you know, building in this area are too focused on the tech and getting things perfect. And I think if you actually focus on an audience and what you really want to build for them, and what can add value to their lives, you approach it in a different way. And I think you start to build in a different way. So like, I don't want to see another mobile game where you're like holding your phone up at the sky. And like shooting spaceships that are flying around you. Like that's been done so many times. It's not fun. But still, you know, developers are making it because it's like, it seems like it's like an obvious application of the tech. But I think you need to look in unobvious ways to unobvious audiences. And you know, the tech will get there. Sam might disagree.

SW I'm actually I'm going to parrot your response. It is not just a content issue, I think it's actually an accessibility issue. In the early stages, like with, even with Laser Draw, a lot of the times you had to you'd spend a lot of time like scanning the room. Yes, that is now been improved by things like the LIDAR systems in like new iPhones and stuff. But it's still not a particularly intuitive action to do, like move your phone around. The problem is that from things like where they are have actually helped a lot where rather than having to download an application to then launch it to then go through the whole scanning process to get an experience that is sometimes not worth it. Because it's again, like Tara said, it's very tech lead and not user experience lead. It just makes things not worth downloading and not interacting with, where they are just made that a lot quicker, because you don't have to download, you can just launch straight from QR codes, and like different launching mechanisms. I think in terms of like, glasses, and the AR glasses and everything that's coming around in those, I think it will still take a long time to adopt, again, with that content, beside of things being a bit of an issue. And I think people will be tied to their phone for a hell of a lot longer.

BP Yeah, what you're saying makes a lot of sense. You know, I remember when Pokemon Go like, you know, swept the world or whatever. And it wasn't like the fidelity of the Pokemon on the ground was what mattered, right? Like that wasn't the thing. It was the social aspect and the you know, the geo aspect and walking around and stuff like that. It wasn't about like, this looks and feels totally real to me. You know, like I'm immersed. That obviously wasn't what it was.

TR Exactly.


BP Alright, let's do our goodbyes, we will shout out the winner of a lifeboat badge. Someone who answered the question with a score of negative three or less, got it up to a score of three or more and 20 on that question of theirs. 'Convert list of integer into comma separated string?' Thanks to Elliott Frisch, we appreciate that. And we'll throw it in the show notes. Say our goodbyes. I am Ben Popper, the Director of Content here at Stack Overflow, you can always find me on Twitter @BenPopper and you can always email us If you liked the show, please do leave a rating and review. It really helps. Tara, let the people know who you are, where you can be found on the internet if you want to be found and where they should go if they want to check out more about Loveshark.

TR Alright, so our website is And you can also get us on Twitter @loveshark_io. If you want to catch up with me my personal Twitter is @taralouisereddy.

SW And I'm Sam and my Twitter is @sjweekes01.

BP And we talked a bunch about Mochi, but you said it's in soft launch, so people can check it on the website. They may not be able to play it or?

TR Yeah, it depends where you are in the world. If you're in the UK or a couple of Asian geographies, you might be lucky. If you're in the states and you really want to get involved, then feel free to message me on Twitter and we can look at getting your test flight.

BP Well, thanks again for coming on. And yeah, love to chat about this space and about the future. It's very fun.

TR Thanks so much, Ben.

SW Thanks for having us.

[outro music]