In this episode of the Knup Sports Show, host Ryan Knuppel interviews Luca Pataky, Senior Vice President for Automated Content at Sportradar. They discuss the use of AI and computer vision in the sports betting industry. Pataky explains that computer vision mimics human vision by interpreting images and collecting data. He emphasizes that AI is meant to enhance products and services, not replace humans. Pataky advises companies to focus on specific problems and use AI to solve them, rather than trying to do everything at once. He also discusses the future of AI at Sportradar, including expanding into other sports and creating new experiences for fans and bettors. Pataky’s favorite sport is Formula One, and he mentions his passion for data in the sport. The episode concludes with Pataky providing his contact information for those interested in reaching out to him.
Ryan Knuppel:
Hey, hey, hey. What’s going on everybody? Ryan Knuppel here, episode 213 of the Knup Sports Show. Thank you so much for tuning in at the time of this recording. I just got back from G2E in Las Vegas and I tell you what, every time I go to Vegas, it wears me out. It plays a toll on me. So I am in catchup mode this week getting a lot done, but it was really nice to meet everybody that watches this show in Vegas. I really appreciate you coming up saying hello and all the meetings that I had there in Vegas, always a great time. G2E, a great conference. So if you haven’t been to that one, I recommend carving it out in your schedule next October. It’s always a fun one to go to. Alright, enough chitchat. We are here to bring on another guest in the iGaming sports betting industry, another leader doing great things, and I’m excited to dive in and understand what all this individual is doing. So without further ado, let’s bring on Luca Pataky. Luca, how are you my friend?
Luka Pataky:
Hey Ryan, I’m very good. Thanks for having me. It’s great to be here.
Ryan Knuppel:
I am excited to have you. It’s my pleasure and I’m really excited to dive in what you’re doing because you’re talking ai, you’re talking futuristic, you’re talking tech, you’re talking my language. That’s the type of stuff that I love to talk about, so I’m excited to dive into that. But first off, Luca, tell us a little bit about you and maybe a little bit about your career path leading up to where you’re today.
Luka Pataky:
Yeah, sure, of course. So yeah, as you introduced me, I’m Luca. So currently I’m a senior vice president for automated content at Sportradar, and yes, I know it sounds really cool and actually it is really cool what we are doing. So we are dealing with some pretty high tech, computer vision, deep learning, automatically creating content, and I’m sure we’ll dive into this a bit later. Me personally, so by education I did master in business administration with major in marketing, but I’ve always liked and loved to try multiple things. I was extremely active teenager, run local newspaper, organized sports tournaments, youth clubs. Also in my career I changed focus quite a bit, did different things. I like to mix things up a bit, keeps me super motivated. I come from Slovenia, so it’s a small country in central Europe. Was born in the capital of the country, Luana, really beautiful city.
I studied there, finished my studies, and then 11 years ago I moved to Austria for my first non-student job, if I can call it like this. And I’m still in Austria till this day living in Vienna currently another beautiful city. And I always had passion for data, I loved analyzing it, I loved taking decisions based on it, drawing conclusions based on it. Somehow always when I kind of did something based on data, got me, gave me goosebumps and I really like it still until today. And I started with Sportradar just under 10 years ago, so a little bit after I moved to Austria. At that time it was fairly small company, just over 600 people today, almost 4,000 truly global company. And I’ve been through a hypergrowth from my personal perspective, but also obviously from company itself and last eight years have been heavily in technology and in innovation actually about eight years ago together with two other senior executives at the time we started innovation team.
We wanted to make sure that we invest time and effort into more long-term future into truly new things and we realized this is going to be fundamental for the future of iGaming, sports betting, sports in general technology. So that’s how it all started. It was a natural step also for the company who was obviously willing to invest significantly, always looking at how to be at the forefront of tech and development in the industry. 2019 was probably really kind of pivotal year in terms of where I’m today. This is where as part of innovation team at the time we started our computer vision journey, we’ve seen that what was once enough in terms of data just wasn’t going to be enough in the future. The fans, the punters, everyone needed more data, became more hungry for insight and actually in order to really differentiate in terms of how you interpret the game and what you deliver in terms of engagement and just fun and excitement, you had to go a lot deeper and we understood that technologies will help us get there and video was always very important for us and it is also very important for our customers.
And of course data is bread and butter of most things we do. So computer vision as the next thing that will help us collect data at scale was kind of a natural step, and that started about four years ago and I’ve been always part of this from the very beginning. We started with a very small team of three people starting to test some things and actually together with experts from within the company grew this from that small team of you could count on one hand to now over 70 AI experts, software engineers, data engineers, deep learning engineers, so very smart people who do a lot of stuff. I kind of come here and talk about it a bit, but there’s clever people behind me who make a lot of great things happen and ultimately trying to reimagine the way our customers and their customers of fans and planters then experienced sports through technology and through data and it’s really exciting. So it all started as an innovation project, the small innovation project which became mature and people that into a full team that’s now actually thinking about multiple sports and how AI deep learning can really change the way we collect data, how much we collect and what we can then do based on it.
Watch This Episode of the Knup Sports Show
Ryan Knuppel:
Wow, that’s a lot there Luca, thank you for that background and everything. And what’s really refreshing to me is hearing that a company like Sportradar, that’s typically when you think of bigger companies, sometimes you think of not wanting to innovate, you think of wanting to stay and not change and kind of do what they do that got them there. What’s interesting to me about Sportradar is kind of taking the opposite approach and really embracing innovation to make sure you’re on the forefront of all of this, and I think that’s really cool. Obviously I’m a supporter of that as well. I just love the innovation side of things. I think if we don’t change with the times we’re going to die. And so I think that’s an amazing strategy that Sportradar has. So AI and computer vision, I guess go a little, we all know Sportradar is at the core a data company, but I want to really want to dive in today more into the AI and computer vision side of things. So maybe give us a real high level overview of computer vision and what that product looks like on your end.
Luka Pataky:
Yeah, so obviously as you already said, so supported there is a pretty big company and there’s a lot of stuff happening. Overall, it’s focused on creating products and services that at the end engage betters like punters and fans. What my team is focused particularly is also one area of ai, which is let’s say at the data collection point. So particularly using computer vision to collect data, if I really try to simplify or put computer vision in a very easy to understand way, what it tries to do is just, it tries to basically mimic what human vision system does. So it takes in images as input and it tries to interpret them. Basically draw conclusion on what is on the images. What that means for us is it actually enables us to collect a lot more data. It’s not about replacing things, it’s about enhancing things and just very, again, to put some numbers behind it, what we see is that just using computer vision to collect more data, we can normally go to anything a hundred times more to what we could collect otherwise.
And this just means that all of a sudden we have all of these information that helps drive products, it helps drive insight, it helps drive odds, it helps drive risk management, it helps actually also integrity. It creates new engagement opportunities. So it really becomes one of the core aspects. But as a technology, it’s part of ai, so it falls under umbrella of artificial intelligence. And for us as a company, artificial intelligence is at the core pretty much everything and we see that’s the future and it drives a lot. It doesn’t just drive the data collection, it drives odds creation, it drives risk management, it drives integrity engagement. It really is becoming more and more of an important piece and asset. It’s about enhancing the experience. It’s about enhancing the products, making them better, deeper data, broader data, and that’s kind of what it is. Current big hit for us, and this is what also we’ve recently came out with was using this technology heavily on table tennis. You may know that table tennis is a very big sports betting sport. It became big obviously during covid, but it stayed quite big. And for us, this is the first sport where we truly, it’s pretty much exclusively AI driven from the point of data collection to how products are delivered to the customer. So there’s very little sort of manual intervention and it’s for the first time where it’s like you could say truly 360 AI from the data collection point all the way to what comes to the customers. So that’s kind of in a nutshell.
Ryan Knuppel:
Table tennis is one of those that every time I hear that comment that table tennis is one of the fastest growing and biggest sports betting, I always think I’m like, I don’t know a single person that’s ever bet on table tennis. And so it’s one of those that the is the data doesn’t lie. There’s data that shows this and this is why data is important as well because my mind thinks nobody betts on table tennis, but the data shows differently. It’s just one of those anomalies that’s very interesting to me when we talk about data and sports
Luka Pataky:
Betting.
Ryan Knuppel:
A little side note there, I like I’ve never bet on table tennis and I love sports betting,
Luka Pataky:
So I mean it is actually amazing, I believe. So for us it is definitely one of the strategic sports, most popular betting sports by turnover. I think in 2022 it’s something like almost 40 billion in betting turnover on just table tennis.
Ryan Knuppel:
Crazy.
Luka Pataky:
And in this year so far we’ve accepted over 13 million betting tickets on table tennis. So it’s a massive actuary sport and you would be surprised how much more you can do. So one of the big things that computer vision enabled us to do is create also micro betts and micro markets and it’s meaning even going a lot deeper into actually live betting. So now ability to bet on how many bounces off the table there will be or whether the ball will touch the net. So all of these things and there is an enormous interest for stuff like that. It’s interesting because at the same time it allows actually people to go into the game and play for a shorter amount of time and then step out of the game. So you don’t anymore need this long-term commitment to the game where you need to sort of predict the outcome of the game. You can really just step in for a short while and be engaged and step out when you have enough. It’s interesting.
Ryan Knuppel:
It is very interesting. Very, very well, let’s switch gears just a little bit. I want to talk a little bit about AI and the human side of it, meaning a lot of times people when they hear the word ai, they get a little intimidated like, oh crap, it’s going to replace everyone in our company and there’s going to be no people left and there’s going to be robots doing everything. How is Sportradar specifically, how are they handling that, I guess false mindset that people might have and are we seeing that? Are we seeing AI replacing people or are we seeing it just creating different types of jobs? Talk a little bit about the human side of that as we go through this whole AI generation.
Luka Pataky:
So I mentioned earlier, so for us it really is about enhancing our products. So it’s not about, I also said earlier that we see that what was once enough is not enough in terms of data. So you still have of course the most basic scoring data and very often this comes from either officials, umpires, and those are facts. And of course also technology is used more and more in these applications that there’s video assistant referees and AI helping with the line calling and all these things. So there is a lot obviously used there, but the basic data is still there and this official scoring data often comes in the same way as it always used to come. This is really about enhancing that, adding those additional layers and allowing us to be more scalable to offer deeper products. And so we look at this as really an enhancement and it just adds on top and it allows us to create a better streaming experience.
So we use a lot of computer vision tech is actually used in creating what we call augmented streaming. So it’s basically augmented reality kind of streaming where we add overlays based on where players are, based on where the action is happening. So it opens up a product that actually is not possible to do any other way, but that way, same way as opening up new betting opportunities with micro betts, there is no way where you can have manual trading on something like number of bouncers in one single rallying table. Tennis really lasts 20 seconds, so by the time you need to settle odds, risk checked for risk, you don’t have time. So these are products that actually really enhance the experience and there you really need automation on the data collection and management. So for us it is not about the replacing, it’s about enhancing and that’s how we look at it and it actually allows us to do probably more things and to do more things better. And so that’s how we particularly look at it in case of sports radar.
Ryan Knuppel:
Wow, so ai, AI is here to stay. We don’t know what the future of AI looks like, but embracing it I think is important. What would you say or what would you suggest to leaders, companies, people in the sports betting egg gaming industry maybe that are overwhelmed by it or maybe just they don’t know how to start, they know that AI’s here, they know it’s here, they know it’s here to stay, but there’s so many tools, there’s so much there. How do you start down that path of just getting into ai?
Luka Pataky:
Yeah, it’s a very interesting, especially now because there is indeed a lot happening and it’s sometimes difficult to pick out the right things to do. We started, I would say still quite early when we started it. And so there’s of course some challenges with this is that there’s not maybe so many reference points and so you build a lot on your own and you need to test and you trial and you fail and you trial again. And that was good for us. I think the way we approached it was there were a lot of fast iterations at the beginning to try to learn what works the best and in what situations AI makes sense and in what situations it doesn’t make sense. So I would always say don’t try to do everything. I think you need to sort of pick a few core aspects of your product or your value proposition and then look at how AI can help you enhance those rather than just maybe trying to look at everything.
Also, every time you of course use technology, you need to kind of use it with care. It’s very easy today to find a lot of tools and you just sort of go with it and then they produce sometimes maybe results that you don’t want, whether you go with automated text generation and it doesn’t quite produce what you wanted, you need to, I don’t think a solution is try to just grab what’s out there and then expect that it’ll work. I think you need to have some sort of iterations. You need to have some trial and error and better focus and then deliver something then trying to do too many things. One other important thing, I think from the very beginning with our journey, we understood that we are not doing AI because of ai. We are doing it because it’s delivering some value for our business.
And so from the very beginning, we stayed very close to the business to understand what the real needs are. I’ve seen many cases where you have great teams, they’re great AI teams and great models delivered, but there’s this challenge of how do they go into production, how do they actually help the product? And you stay somewhere in between, I think from day one, define those together with your product teams together with the rest of the business and really focus because sometimes you can reach great things with ai actually not by needing to invest for one and a half year, you can actually get there really quickly if you focus enough and you really know what your objective is. So I know it’s not quite the best sort of recipe maybe, but I think it’s a little bit of a learning from our own journey and I think it helps if you stay focused and business oriented.
Ryan Knuppel:
No, Luke, I think that was really, really wise advice. Find the problem, focus on the problem, what’s the problem we’re trying to solve here? And then figure out how AI can help ease that problem. It’s not going to be the end all, be all, solve everything, but figure out how you can start to fix it, how you can start to make it easier. I think that was really, really wise advice there.
Luka Pataky:
We’ve sometimes also seen that even we went to prototyping and we did a small prototype and actually it already delivered quite a lot of value and it was maybe just a month of work, and then we went from there and in iterations we build on it. But even just something that we thought would just be a prototype and we’ll throw it away, we actually could use it and it showed value and that was because the goal objectives was really clear and business really knew what they want to get from us. And prototype already was quite good.
Ryan Knuppel:
Yeah, there’s so much, and I’ll use the word noise lightly here. There’s so much noise around ai, there’s a lot going on. And so it’s very easy to be like, Ooh, it can do this. And then, oh, ooh, it can do this and it can do this. And then all of a sudden you’re like, well, crap, I’m just overwhelmed and I’m not really doing anything. I’m just trying all this stuff. And so I really like how you said focus on a problem and attack that, attack that problem. Kind of go deep on one area and then figure out how to go deep on another area if you need to. But I really like that. So Luca, okay, what’s the future look? So I feel like we’re already into the future. We’re talking ai, but more specifically for you, Sportradar computer vision, what’s the future look like for you guys? What are you doing in the future to continue to develop this product?
Luka Pataky:
So the future is definitely still, and I know we keep repeating this word, but it is still going to be very AI driven. Fortunately, I guess also for me and my team, that’s good news. Yeah, so as I said earlier, it’s going to be part of every product and service we offer where it’s going to make it better for all stakeholders. We, for example, a big part of our business is also integrity. We work together with leaks very closely to make sure sports are free of match fixing. And that service is more and more AI driven. You can imagine a lot of data comes in, it’s impossible to actually process it fast enough any other way, but AI way. So AI will be part of everything in terms of particularly automated content now, as I mentioned, launched it on table tennis where we are powering actually more and more products.
So not just the sort of data collection in general, which powers all the betting products and visualizations. We are launching, as I said, micro betting product. We are launching augmented streaming product. All is actually completely AI driven. And we are looking also to launch more products based on those capabilities. And we are looking at other sports. So we are already pretty good with tennis, we are also looking at basketball. So for us it’s like how can we take what we’ve now built and launch for table tennis and how can we take this and do actually similar experiences for fans and punters and operators for other sports?
Ryan Knuppel:
There’s
Luka Pataky:
Definitely interest. We see a lot of demand and those experiences are creating also a lot more engagement on operator side. So there’s definitely demand. And yeah, the future for me and my team is how do we expand this in multiple sports? How do we help generate more value from sports we offer by actually putting AI on top and collecting more data and creating new experiences. And I can see definitely for still a couple of years in the future, there’s going to be a lot going around this and we will have to continue to invest and we’ll invest to make sure that the products we put out to market are the right for what the market needs at the moment.
Ryan Knuppel:
Amazing, amazing, Luca. Well I really appreciate your time here today. I know you’re a busy man, so we’ll let you go here. But on a more personal level, what’s your favorite sport? What’s your favorite sport just to be a part of?
Luka Pataky:
Actually, I like Formula One, which is, I like motor sports actually, but not the two wheelers, more the four wheelers. I also like to drive a lot, so I like, but four wheelers not,
Ryan Knuppel:
I’ll tell you what, I like F one as well, but I didn’t like the thought of F one this last weekend in Vegas because it’s coming to Vegas here in a couple weeks. F one is, and it was like a nuthouse of just construction. I mean they’re going to run F one down the strip of Vegas. That’s going to be absolutely amazing site. I wish I was there
Luka Pataky:
And maybe the reason why I like it is also because also a little bit obsessed with data in F one. So I kind of like this. I can’t get bit of this passion of mine, I guess.
Ryan Knuppel:
Awesome. Awesome. Well, Luca, any last words? Anything we missed? Anything you wanted to say before we get off here?
Luka Pataky:
I think not. I think we covered most things. So Ryan, thanks again for having me. It was great to be here and hopefully also lots of interesting thoughts for your listeners and yeah, thank you.
Ryan Knuppel:
Hey, how can somebody get ahold of you or your team or check into more of what you guys are doing? We’ll put the links in the show notes, but maybe shout out how they can get ahold of some of you.
Luka Pataky:
Yeah, you can reach out to me on LinkedIn. I’m there pretty active, so if you want to talk to me more, I’m available there. Otherwise, yeah, we keep trying to make ourselves visible in different conferences. I’m sure we’ll have something interesting in the future as well. But yeah, reach out to me on LinkedIn.
Ryan Knuppel:
Awesome, Luca, have a great day. Thank you. And I look forward to seeing your success with Computer Vision and we’re all at sportradar.
Luka Pataky:
Thank you, Ryan.
Ryan Knuppel:
Thank you. Have a good one. Alright, that was Luka Pataky of Sportradar while doing some amazing things in ai. Some really wise advice there for some of you trying to dive into ai. I know on our end we’ve invested heavily in our coded content products, so we’ve been really putting out a lot of AI content across all the major US sports, and that has not been easy. It’s a lot. You have to dive in, you have to focus on the problem and really attack it. And so I urge you and whatever you’re doing in business to take that advice that Luca gave and execute and just start going down that path and learning. It’s a lot of fun once you do start learning. So, all right, that’s it. I’m Ryan Knuppel, K N U P on all the social media channels. I really thank you for your time listening to this show. That’s episode two 13. Looking forward to many more to come. But until next time, take care. Stay safe. We’ll talk to you soon. Bye-bye.
Relevant Guest Links
- Learn more about Sportradar on their website.
- You can connect with Luka Pataky on LinkedIn.
More Knup Sports Links
- I refer to Knup Sports often in my show. That’s our sports passion site. Check out KnupSports.com here.
- If you are looking for sports content services you can find more information on the Knup Solutions website.
- Listen to more episodes of the Knup Sports Show.
- Want to be a guest on our show? Setup a time that works here.
- Subscribe to the Knup Sports Show on iTunes here.
