Behind the scenes of the Bundesliga - AWS partnership

0:04Hello everyone and welcome back to another edition of Streamtime Sports. I'm your host Nick Meacham doing a very special episode of the podcast here in Düsseldorf at the Mercure Spiel Arena for Sports Innovation 2026. I've been talking to lots of people across the event and on stage, but you're going to hear a little bit more from one of the most important companies, perhaps, in the sports and sports media landscape. And that is AWS and what they're doing both with the Bundesliga and also with global sports properties like the NFL, F1, PGA Tour and loads more. In this episode, you're going to hear from two different people.

0:37You're going to hear from Joanna Prestel, who is the Global Sports Partnership Lead from AWS, talking specifically about the details of the Bundesliga relationship. and also from Julie Souza, who is the Global Head of Sports of AWS, talking more widely about the relationships they have with all sorts of major sports properties like the NFL, F1, PGA Tour, and so on. So I'm here with Joanna Prestel, who is the Global Sports Partnership Lead at AWS. And AWS's major partnership when it comes to sports, really, at least from my view, is the Bundesliga relationship. And she's going to give us a bit of an insight into the dynamics of that partnership.

1:12So, Joanna, welcome to, I guess, Sports Innovation and also to Sports Pro. Take us a little bit of a look into the relationship, the partnership that has existed for quite a while now. What are some of the particulars and specifics about that relationship that can set the scene for how AWS works with the Bundesliga? Yeah, I would love to. First of all, thank you so much for having me. Super excited to be here at the event and talking to you about our partnership. Yeah, AWS and Bundesliga have been partners since 2020. and what I really love about this partnership is that it's a technology partnership, right?

1:46But we don't develop technology for technology's sake. So we really always have the same vision of innovation and really always working backwards from the fan. What is the challenge of the fan or how can we add value for the fan? And I think that's what makes this partnership very special and that's also representative with the flagship use cases we developed together and yeah happy to talk about those but yeah that's basically I think it's sometimes lost on people about the fact that the Bundesliga takes a lot of the responsibility of that glass-to-glass distribution yeah and that is obviously very different to a lot of the other European football leagues and indeed most of the football leagues right across the world you obviously help a lot with enabling that so talk us through some of the specifics of the partnership that you are enabling basically the Bundesliga to achieve with their technology and broadcast yeah 100 % because they are so independent, we're actually able to help them innovate much more than other leagues, right?

2:44And yeah, so first of all, what we did when we started the partnership is really the foundation, right? The data. Like how can we look at these 150 ,000 hours of 60 years of Bundesliga footage, video footage, right? And how can we put that in the cloud and make sense of it? And how can we, yeah, create flagship use cases like Bundesliga match facts So that is actually one, the first flagship use case we created, where we analyze 3.6 million data points per game and offer that to the fan during the live match in the broadcast, but also in the app. And that's where the value comes back, right? Because it's not just a fact that we can look at, but it really creates context for the fan.

3:31So I don't know if you know the Bundesliga match facts. Yeah, I think we've talked about a little bit on the pop. but give us a bit of a, could you just talk us through what it is? Yes. And the nuance between when you say, I guess, we in that, like what is the dynamic when you say we, is it in collaboration? Is AWS playing the role in looking at that data? Or is it more that you're just helping Bundesliga have access to that data? Yeah, so it's really both. So we have 16 to 20 cameras in each Bundesliga stadium. And within one second, we analyze 25 points per player. and we capture that data and thanks to AWS serverless architecture that is captured in the Bundesliga data hub and within under one second we analyze that with data analytics, machine learning and make sense of the data, right?

4:19And that's how we collaborate and then integrate that into the live feed and into the Bundesliga app. And then for instance, I mean, I think the easiest example is if you have Harry Kane scoring a goal and it's just a likelihood of like the goal probability was like two percent and that's something where a fan is like wow that makes that makes that moment special right like you think about wow this is that's very very low and that gives the fan context and value and the same goes for um like giving context throughout the season not just saying hey this was the fastest player in that moment of the game but actually of the entire season and that again um creates a lot of value for the fan to interact with the live game.

5:04So you're helping them get access to a lot more data than they ever have. And the thing that the match fact is a way of, I guess, translating the more the nerdy data and metadata and the like that is created by that and translating it into fan speak. Is that how you would describe a lot of basically the enablement of what, when you have all the access to the information? A hundred percent. I think it's, it goes both ways, right? because we have Bundesliga match facts that are integrated in the broadcast that are more like the easy effects to understand for people who might not be as data-driven or as tech-savvy, but still are interested in these numbers.

5:39And then we have match facts that go a little bit deeper where fans can dive into in more detail in the app or on the website. And I think we serve all kinds of fans in that way, right? Absolutely. So just curious, talk us through some of the other examples of where AWS is, I guess, bringing that technology to the fore? Yeah, so, I mean, one big challenge for all sports leagues is content localization, right? Like, if you think about Bundesliga, they have one billion fans worldwide, five or nine million actually in Asia, but currently the game is only produced in English and German. So, however, the leagues, they're not big organizations, right?

6:20They cannot just, like, add resources. So that's really where our work comes in as well. Like how can AI help to localize content immediately, live game, post game, etc. So we do have these content. One project we have is called ContentOS, where we automatically translate content from live game and post game for these 75 plus partners Bundesliga works with internationally to really give access to the game for fans that usually wouldn't. so what thing that things that is difficult for people to get their head around particularly around sports and partnerships is where the lines draw and who does what yeah i see aws is like the i don't know the roads the the architecture the infrastructure underneath that's enabling that's true this league to do the work uh what are you what are you doing and what are you not doing in terms of just for people to understand the delineation there of bundesliga's role in creating all that are they coming to you with a problem and saying and you give them a guide of like all right here's how you build it on our infrastructure or how much are you having to take actually the ownership of developing that yeah so we do most of the development but what we always i mean we are doing everything collaboration right so what we usually do is we work backwards from the fan we work backwards from from our customer and we create these uh prfqs so really these press releases of products like what is the end result we want to achieve And then we work backwards from that solution.

7:49Say, okay, what is actually the technology we need for that, right? And that is something that how we work at AWS with our partners and we work with Bundesliga in that way as well. And then it really depends per project, to be honest. I think one really sexy project that we're very proud of is also Bundesliga Stories, for instance. It's another content scale example where Bundesliga creates around like 4 ,000 articles per season, like long-form articles. But nowadays, fans require these short-form snackable content, right? And there again, Bundesliga doesn't have the resources to expand their team.

8:27So we work with our Amazon Bedrock, Amazon Recognition, right? We analyze these 4 ,000 articles and we create them in like snackable AI stories, right? and that's something that is in the Bundesliga app. They create 400 stories a month in the app with that, with amazing results. So that's one great example. That is a great example. It feels like the whole industry is really focused on content creation and management now. It's one focus, right? It's one focus, but it's a pretty big one because everyone's trying to engage audiences. Correct. Can you give any insight as to how automated is that? Are the Bundesliga having to still manage it with people or is it able to be fully automated?

9:11It's almost fully automated. It could be fully automated, but Bundesliga wants to have that final check because it's their content in the Bundesliga app, which is their biggest outlet to fans and access to fans. So, yeah, this whole process from the article screening, identifying the right image of the player of that game with the right title and creating that snackable piece of content is done automatic in under one minute. And then the Bundesliga editor can look at it and say, okay, final goal. Let's put in the app. And final question, is there any other examples of where AWS is being used by the Bundesliga that you'd want to share and point out as an example of the partnership?

9:55Yeah, we have something in the making right now, which is really exciting. We're launching that for the World Cup. it's captain it's called an it's like an AI companion in the Bundesliga app because we have seen again working backwards from the fan to that fans are juggling during a live game like they look in chats they look at the stats and then they see the live game but it's really fragmented the experience so we all put that in one place in the Bundesliga app where you can ask all kind of statistical questions where the AI companion is really helping you to understand the moment the match, the situation.

10:29And that's perfectly for the World Cup, also for new fans who don't know football that well. So that's something that we're really excited about. Yeah. I mean, for those that are following the journey of different European football leagues, the Bundesliga is really the pioneer in many aspects around technology and broadcast production and content. So big thank you, Johanna, for sharing some more insight around the AWS component of that and a little insight into how it all works. We've just heard from Johanna about the Bundesliga relationship. Now we hear from Julie Souza about AWS more widely right after this.

11:06What's happening now and next in sports tech? As AI adoption accelerates rapidly, the inaugural Global Sports Tech Report from Sports Pro and Sport Radar reveals which technologies and trends are driving real business impact across six core areas of sport. Informed by a global survey of senior executives and insights from leading tech and sport organizations, the report offers a 360-degree view of what's working today and where the industry is headed tomorrow. Access the full report via the SportsPro website.

11:44I'm joined by Julie Souza, who is the global head of sports for AWS. So we're going to take a little bit of an insight into the various, and there's plenty of them, examples of the relationships that AWS has with different sports properties right across the industry. So, Julie, welcome to Sports Pro. Welcome to the Sports Innovation event. How have you found it so far? I love this conference. I love it so much. We get to see technology and innovation in action. So it's one of my favorites. A lot going on out there. That is for sure. Now, I want to take a look at some of the examples, as I mentioned earlier, that you're seeing, that you're involved with across the different sports, because it feels like AWS is kind of everywhere, pulling a lot of strings for a lot of different sports properties, whether it be international or US-based.

12:26So can you share some examples and perhaps learning from those relationships you have? Yeah. So we're fortunate to work with some really fantastic global sports properties, Bundesliga being one of them and why we're here. We work with the PGA Tour, with Formula One, the NFL, the NBA. It's an impressive and distinguished list that we are privileged to work with. And the use cases vary, right? We like to say we work backwards from our customers' problems or challenges. So whether it's the NFL with player health and safety or it's driving fan engagement through analytics or other means, sort of up for that challenge and really try to position ourselves not as a technology vendor, but as an innovation partner to those leagues.

13:10So there's a lot of major players you mentioned already off the cusp. They all need different things. Sometimes you've taken a partnership sponsor-led approach where you're really enabling the visibility of the AWS brand. Others are very behind the scenes, pulling a lot of strings, giving them the infrastructure to make things happen so just give us some specific examples of some of those things particularly the u.s examples you shared of how your what are some specific i guess products or outcomes they're trying to achieve by working with you also so i'll start with maybe the nfl um because there are a lot of different things we work on with them one um is their analytics platform next gen stats so those are analytics that drive understanding and engagement with fans and that's you know pulling data off of the field and off of the players to be able to create those insights.

14:05But also, it's a high-contact sport with a lot of injuries. So another challenge that we have tried to address with them is around player health and safety and capturing, again, even a more granular level of data to help inform player health and safety initiatives. And I think I would love to dive a little deeper on that one. And because it was such a pronounced business problem, right, we have high rates of concussion, lower extremity injury rates. How do we lower those rates? And with the data that's being captured, and I think so many of these use cases actually come down to how do you capture and analyze data to make decisions, right?

14:47pulling this data in and being able to identify plays that were more injurious and watching the NFL actually change rules in its game to address that. It was remarkable. Two years ago, there was a kickoff, which happens, and they found that that play caused two times the injury rate and four times the concussion rate over a normal pass or run play. Also, they found that the kickoff was not actually being returned into an exciting play. So it was a boring play that caused a lot of injuries. So they decided to use machine learning and all of this data to simulate 10 ,000 seasons worth of data to optimize for lower injury rates and a higher return rate.

15:30And they changed the rules in American football based on that data. And what they saw was exactly what that sort of machine learning modeled out. The injury rates normalized. and last season they had 74.5 % greater returns. So a more exciting game as a result, while obviously making it a safer play for athletes. So I just think examples like that where we can tangibly point to like real value being driven off of the use of data in really strategic ways, that's super exciting. That's a pretty incredible example, Nishan. For those that maybe aren't NFL fans, and I'm one who does watch quite a lot of it, has been quite a change to the way the kickoffs and the dynamic of that experience has been delivered now.

16:16When, how far can you, I don't know how you can, where you can take us here, but to do that modeling, what goes into that? Is it you guys hand over a tool or a laptop and say, all right, here's all the, it's all set up now to model and go for it? Or are they giving you some criteria and you're chucking this AI led ML machine that does all the dirty work and comes back and spits out a big number? Just talk us through how that outcome actually happens. Because you hear simulation, and I get the thought of video game simulations where you see something play out. Is that kind of, how far does it go?

16:49Well, and I think, again, it's interesting to think about, like, it goes back to, first of all, the data capture, right? So it used to be that, and they still do, capture data from chipping the players. So each of the shoulder pads and the football itself are chipped. And so that's collecting data. But they've also installed cameras in the venues and they're collecting a much more granular, about 29 points on the body every 60 seconds is the amount of data that's being pulled off of the athletes in the game. Right. And so, first of all, it's organizing and storing all of that data. And then, yes, you're developing models to help simulate different things.

17:24And it's a very collaborative process. The NFL is involved with their data scientists and their partners in that space. We obviously have deep understanding and proficiency and expertise in machine learning and models and simulation tools and sort of the predictive nature of these things. So it's a very collaborative process. And that's what makes it work, actually. It's not just we'll take it away and come back to you with an answer. And it's also not just give us your tools and leave us alone. It's very much how do we work together knowing what the objective is we're trying to solve for. So when I hear the term data, I think a lot of people will think about traditional statistics that are created, maybe some of the betting and data agencies that service and create content and data around that.

18:09What's the difference? When you talk about data, it's definitely not the same type of conversation we're having there. Can you just talk about the differences between those different buckets or pools of data that you're capturing versus maybe the traditional data sets that the industry and consumers might be seeing? Yeah. So first of all, data, I usually use that to mean all content, really. All content is data. Video is data. You know, analytics are data. You know, I always say when I started off as a young sports fan, the raw data of sports was a box score. And it's exactly what you're describing, that which people can see and observe, right?

18:45You know, completions or goals or things that are very obvious. the level of granularity we get off of tracking data now, it's that raw data is X's and Y's and Z's. It's literally like degrees of latitude and longitude. So if you're looking at it as a fan, it makes no sense, right? It actually requires the application of machine learning to be able to understand those analytics or those data points in relation to one another and then contextualize them. So it's a much deeper level now. And it can sound a little off-putting. I think sometimes like, oh gosh, all this data, it's very heady, whatever.

19:25Everybody's sort of doing that now and it's getting easier and easier to capture this data. And I think it's the expectation of fans, truthfully now. People expect to have insights that help them understand the game better. And I think it's really important for leagues to embrace it because you're always sort of getting the next generation of fans, right? And so that means two things. One, speaking to them the way they want to be spoken to and the next generation of fans, they are tech savvy. They are data literate. But also, you know, how do you use this to edify an audience and teach them about the sport so that it doesn't feel like, oh gosh, I don't want to get involved because I don't know the rules or I, you know, if you can make fans feel smarter about what they're watching, they'll watch more of it.

20:04They sure will. And I've noticed AWS's data points being shared in a lot of major broadcasts in F1 and in the NFL and the like. And normally there are new data points that didn't actually exist until now, until AWS has been involved. Outside of the, let's call them the more sexier points, which is data, which I'm not sure. It sounds like a counterintuitive line. I think data's sexy. What are some of the maybe less obvious things that might be standing out that are actually having a major business impact? Because I think of AWS is, again, that infrastructure part that is quite transformative to an organization if you get that right, right?

20:41I don't know if you're trying to work out a metaphor here, but I can't really work out what the best way to describe that is. So what are some of those things that might not get the coverage on the broadcast, but have been really influential, I suppose, in impacting some of those sports properties you mentioned earlier? Yeah. I mean, I think, you know, it is that infrastructure layer that allows you to really organize all of your data, which again is any sort of content. And we've been talking a lot about the data that comes out of sport, But equally, there's a tremendous amount of data about sports fans, you know, and they're transacting and demonstrating their fandom in a lot of different ways.

21:13And so how do leagues or properties or teams, clubs capture that data and develop insights from it so that they can speak more directly in a more personalized way? Right. And I think that's sort of this next frontier. People expect more personalization and they're out of their content now. And so how does how does sports rise to that occasion and how do you speak to a fan? And, you know, Julia as a sports fan is different than Nick as a sports fan. And so how do we speak to each other in ways that lets each, you know, fan feel seen, heard, valued? And when you can do that and create that connection and make a fan feel valued, you're going to be a more avid fan.

21:51So is that those sorts of data points that you're referring to, obviously, like things around consumption in particular? Sure. Ticketing, e-commerce, you know, merchandise, subscription products, you know, betting, fantasy. There's so many different touch points for fans to express their fandom. How do you pull that all together? We had a property that had, actually, well, it was the NFL. They had 90 billion rows of fan data, but really only visibility into 12 million fans. So they had all of this disparate, fractured, duplicative data. And they really went through this exercise of structuring that data and creating a single unified view of each fan.

22:29And when they went through that exercise, visibility from 12 million went to 70 million. And with 70 million fans, they were able to do a little bit more persona based marketing. And they saw two to three times of increase in open rates and things like that. So you've got more engagement as you started to get more personalized, even on a persona basis. Now, imagine that at a hyper personalized one to one basis. You're going to see real tangible business results from that. For sure. So NFL is obviously the best of the best, the biggest, depending on which metrics you use in terms of sports properties.

23:02And that's 90 billion data points is a lot of data points. For those that are maybe challenger sports, the next tier beneath or a couple of tiers beneath those major properties that we touched upon earlier. The idea of getting their data in order is I think for most business people now is understood as a must do. Yeah, it's quite an overwhelming kind of thought process to try and get aligned with all the different platforms we've got. Talking firsthand experience here of somebody who runs a media company, I've been wanting to get all my data in place for one time, and it's been easier said than done.

23:35Just talk us through how is AWS able to therefore make that happen for those small properties? Because some people will just go, oh, well, the NFL is doing it, but I'm not the NFL, right? No, I mean, I think it's very accessible now because you used to have to invest in infrastructure that sometimes, you know, server racks and things like that, this capital expense that made it sort of hard, especially what are we going to do with this data? How are we going to monetize this data? It made it a little bit speculative for them. I think you're right. People understand the value of data now, but still that capital expense makes it, you know, you're making decisions there.

24:09Now with cloud services like AWS, you're able to turn, you know, CapEx to OpEx. So you're only using, you know, only paying for what you use when you use it. And so it's not this always on thing that's draining. It is it's become much more economical. And then you obviously can avail yourself of resources to help you understand, well, what do we do with this data? How do we exploit it to, you know, drive business value? So it's I think that's just I think people. And I do think this allows some of those challenger brands you're talking about to leapfrog. You know, those that really lean in and get to these super personalized experiences.

24:46I think they have an opportunity to sort of carve a really powerful niche for themselves. So you get approached by a challenger sport who's looking to take the next step, could be older, could be newer as well. There's different case studies or reasons why someone would look at these types of things. that notion of like trying to get the justification to start committing to investment in their infrastructure their systems is sometimes it sounds obvious sometimes it's really hard to get buy-in from your stakeholders investors owners to do so is there anything you're able to do to help them navigate that challenge because obviously it's in your best interest for them to understand and be able to get that commitment as well I think it goes back it's a nice full circle to the where we started the conversation, which is, you know, what's, what challenges are they trying to solve?

25:31Right. Um, oftentimes it's, yes, we don't really have a handle on our fan data. Well, if you can't don't have a handle on your fan data, then you can't really speak to them in a way that's going to get them to act the way you want them to. Right. So it's, it's always going back to what's the business challenge we're trying to solve here. Um, and so we like to start there. We don't start with technology. I never do. I start with what business problem are we trying to solve? And then is technology the right answer? What technology? And certainly the answer is generally yes. Technology can play a role here.

26:01And then how do we help? How do we help you do that? But I've said repeatedly, technology for technology's sake is not interesting. Technology that solves a business problem, that drives fan engagement, that drives operational efficiency, that drives revenue opportunities. That's interesting. So what are some of the main areas that are those business challenges that you're getting approached about? I'm guessing it's about operational efficiency. I'm guessing it's a lot to do with media content creation management. Are they really the main? And of course, just generally sorting out the data side of things.

26:30Are there anything specifically that you would point to as like either maybe trends that are getting focused on right now or just topics that you're talking more and more about that are more of a priority? Yeah, I think we're seeing a lot. I mean, personalization is a big trend. Globalization is a big trend in sport. And we see that. I mean, this is a very global conference where we are right now, right? So I think those are trends that call for technology and support in that area. I think you touched on content creation, and I'll speak very quickly to an example of PGA Tour, though Bundesliga has done a lot in this space as well.