How ELF reduced their carbon footprint by broadcasting on AWS

Good morning, everyone. Welcome to re:Invent on our first day. I'm glad we get to spend it talking about the intersection of media and entertainment and sustainability, which is my favorite talk topic.

So I'm glad I get to do it here. In order to put the mindset for this session in your head, I've got three questions:

1. How many of you feel uncertain about the impacts that climate change is gonna have on our society?

2. How many of you would like to be able to feel more confident in speaking about how sustainability relates to media and the media industry?

3. How many of you would like to be able to say or feel that you're doing your part in order to make a difference?

With that mindset, hello, my name is Jason O'Malley. I'm a senior partner solutions architect here at AWS and I'm a member of our environmental sustainability and media and entertainment technical field communities. We also have a media and entertainment sustainability working group that I'm a core member of.

So I'm really excited to be here. I'm also really excited that we have the European League of Football and Giacomo Colombo, the CEO, here today. They have an amazing story. They are an American style football league that's based in Europe. They started just three years ago and this year, they made the bold decision to switch to a model in which they produced the majority of their games using a remote live cloud production workflow.

So I'm really excited to talk about that, especially because this workflow was actually recognized by Sports Video Group Europe earlier this month for a Sports Broadcasting Achievement Award. So we're gonna deep dive into the architecture.

I've also got my colleague, Felix, here. Felix is a solutions architect that has been working with the European League of Football this entire year as they've gone on this journey.

So without further ado, let's dive in. Our session today - first part, I'm gonna cover the topic of sustainability as it relates to AWS. Then we'll talk a little bit about how does sustainability relate to the media and entertainment industry.

Then I'll bring those back together and talk about how does that relate to live cloud production. Giacomo will come up and talk about the founding story of the European League of Football and talk about how they came to the decision to migrate their workloads to doing the live cloud production on AWS this season.

Finally, Felix will do a deep dive into the AWS architecture that made this achievement possible.

So we're here today, we're at re:Invent and we're talking about the critical issue of climate change. Now, many of us have seen the headlines, some people may have read the reports, some people may have seen some of the impacts of climate change firsthand.

What becomes apparent when you read the reports is that in order to attain the type of carbon reductions that will be necessary in order to mitigate the worst impacts of climate change, it's gonna require both cross government but then also cross industry effort in order to find inventive ways to rethink the way we do our businesses' usual processes across the board and remove carbon footprint from our processes.

One of the ways that Amazon is thinking and working about the issue of climate change is with the Climate Pledge. So Amazon co-founded and was the first signatory of the Climate Pledge back in September 2019. This is a pledge to reach net zero carbon emissions by 2040 which is 10 years ahead of the Paris Agreement.

One of the things Amazon does as well is work with other companies that would like to achieve the same goal and welcome them in. To date, we've had over 400 companies, some of those even in media and entertainment, that have joined and signed the Climate Pledge.

The only areas of action that a company would need to agree to are regular reporting, carbon elimination and credible offsets. So if these sound like goals that align with your organization, you can find out more information about joining on the Climate Pledge website and we'll have a QR code for that at the end of the session.

Now we'll talk about media and entertainment and the carbon footprint in the media and entertainment industry. Overall, an increasing number of media and entertainment companies have begun reporting on their carbon emissions in some form - their greenhouse gas or GHG emissions.

This increase in reporting has also driven an increased awareness about what's actually in the content of those reports, and then also what's the environmental impact of an organization's carbon footprint?

Doing a quick recap of GHG protocol - GHG protocol is the predominant global standard for reporting GHG emissions. If you're reporting GHG emissions, you're going to report in Scope 1, 2 or 3.

Scope 1 emissions are gonna be emissions that are direct emissions from the organization that are owned and controlled by the organization. An example would be in media - if you have an outside broadcast truck and the emissions coming out of that tailpipe, that would be Scope 1. Or the natural gas that's utilized in order to heat your owned and operated facility.

Scope 2 emissions are emissions that come from indirect emissions from the purchase of electricity, steam, heat or cooling. Now, in the context of media, this is primarily gonna be electricity to power things like our cameras, our lights and our server infrastructure.

Finally, Scope 3 - these again are indirect emissions and these are from assets and activities that are not directly controlled by the reporting organization, but that the organization indirectly impacts as part of its value chain. This can relate to goods and services that the company purchases.

One that I'm gonna call out specifically for this session is that Scope 3 will include business travel. And so that'll be related and we'll think back to that later.

When we were preparing for this talk, we were trying to think about what's the current state of the media industry. So we did a relatively simple exercise where we took 10 significant media companies, downloaded the reports and then averaged out that percentage of Scope 1, Scope 2 and Scope 3 as a percentage of total.

What we found when we averaged that out is that across those 10 companies, Scope 1 was around 4% - again, those are those direct emissions. Scope 2 on average was about 5% again, that's from primarily purchase electricity. Then Scope 3, which again includes things such as business travel, was 91%.

So if we're looking for where within media is the area in which we can target for the biggest impact, Scope 3 is gonna be a prime target. This aligns within the media industry as well.

Some people here may be familiar with Albert. Albert is a group that focuses on the environmental sustainability of film and TV production. They produce reports and one of the reports that they produced in the Albert Annual Review of 2021 found that travel remains the largest part of a production's carbon footprint.

The report then goes on to say that if you're able to reduce travel or change travel of any kind, that will also help impact or reduce your overall carbon footprint.

It's important to think about this on two levels - on one level, you can think about it on your individual production or organization level, but it's also important to think about it collectively as an industry. If this is the largest part of a production's carbon footprint, then what can we as an industry do in order to reduce our impact?

I'll give you a bit of a comparison. A lot of folks that work in production will be familiar with the concept of a production budget, which is the total amount of money that you can spend within and you don't want to go over that budget.

Well, within environmental sustainability, scientists have come up with the concept of a carbon budget. A carbon budget is the total amount of carbon dioxide equivalent that can be emitted into the atmosphere until scientists predict we will hit those 1.5 or 2 degree temperature rises and receive even worse impacts from climate change.

In short, any meaningful reduction of our carbon footprint will help impact that bigger issue.

Now, moving into talking about how do we at AWS think about sustainability - we break it down into three categories:

1. Sustainability of the cloud
2. Sustainability in the cloud
3. Sustainability through the cloud

Sustainability of the cloud relates to the sustainability efforts of our data centers and cloud operations. What this comes down to is that AWS scale allows us to achieve a higher resource utilization than is gonna be possible than the average on premises data center.

International research firm 451 Research, which is part of S&P Global Intelligence, has run multiple reports across multiple different geographies. What they found is that when you migrate a workload from on premises to AWS, you can save a carbon footprint of nearly 80%.

It also found that within the US, AWS is 3.6 times more energy efficient than the median enterprise data center surveyed. That number is five times more energy efficient for the average enterprise data center surveyed in Europe.

This also relates to how Amazon is the largest corporate purchaser of renewable energy globally. We have a goal at Amazon to be powering our operations with 100% renewable energy by 2030 and we're on pace to hit that goal five years early by 2025.

Related back to AWS - as of 2022, we have 19 AWS regions that are powered with energy that's attributable to 100% renewable energy. That includes all of the available regions within the US, Canada and Europe.

Then, sustainability in the cloud is a continuous effort that's really focused on the energy efficiency and carbon footprint reduction of your actual AWS environment. This is a continuous effort.

One of the resources we have that supports our customers with this is our Well-Architected Framework for Sustainability and the Sustainability pillar. This pillar is a white paper that contains best practices, operational guidance and design principles that help you optimize and achieve your sustainability goals on AWS.

A couple of the best practices that we can borrow and look through the lens of media are:

• Right sizing your workload - say you have a video mixer workload running on an Amazon EC2 instance. You right size it, find out what are those computational requirements needed to run that application successfully and you right size it so you're not wasting extra CPU cycles.

• Another example could be if you're producing live sports events on the weekend and you have on demand resources, then just shutting those resources off during the middle of the week can also help reduce your AWS carbon footprint.

Finally, sustainability through the cloud - this is different. This is when you actually utilize AWS to transform and achieve an outcome related to sustainability that's outside of the cloud. This is where we're gonna be focusing our section here today.

Diving in - sustainability through the cloud. Emphasizing that the first two that I mentioned (of and in the cloud) are really about optimizing the IT efficiency, the energy efficiency of those components. Sustainability through the cloud allows you to transform the way you do your business as usual operations.

We're gonna take a look at a use case of a through the cloud use case which is live cloud production, which allows you to produce live events through the cloud.

Before doing that, a level set on live production - high level view of it. If we take sports as an example with live production, you're gonna have your cameras, you've got a stadium, you've got your video and audio feeds and then those route back to components like a video mixer, audio mixer, graphics, replay machines.

Typically the legacy way of doing this would be those would commonly be housed on site within an outside broadcast truck or an OB truck (which you see in the bottom corner there). An OB truck, as you can see, it's a large truck. The reason it's large is it not only has to house that specialized video equipment like the video mixer and audio mixer, but also the operators of that equipment as well.

Now we'll shift - if we shift into live cloud production, you'll see it's got the same functional components. We still have our video mixer, audio mixer, graphics and replay machine, but they're virtualized, running on software on an Amazon EC2 instance, for example. But the same functional components are there.

What the difference is - so now you have your video mixer, say it's running on an Amazon EC2 instance in an AWS data center. Well now, you have an operator instead of being there on site controlling it, they're remotely controlling that machine, but the machine is in the AWS data center.

This can happen in a couple of ways - in the case of the European League of Football, they actually have the group co-located together in the same facility all remotely producing the game. But it also opens up the possibility that those operators could work from the comfort of their home, which is something that our industry saw a lot of, especially during the most challenging times of the COVID-19 pandemic.

And because this can be a little abstract, there was a time when I uh hadn't done this before, I actually did it in real life. And because it's a bit extract, you may have a vision of, I, I understand how live production works. I've seen that. I've known that for years or decades, but the concept of live cloud production might be a bit abstract.

And so what I've done is we pulled together a few behind the scenes examples from the European League of Football. And so what you'll see is you're gonna see those same similar hallmarks, you're gonna see the multi view your video, your audio mixer, those are all still functionally there.

One thing that is different is if you look at the bottom right hand corner and you look at the transporter van. So that's an example of the transporter van that the European League of Football utilizes in order to bring the still required gear to site. What you notice about it is it's smaller and one of the reasons it's smaller is that you no longer need to transport the video mixers, the audio mixers to site and the operators that would normally work out of the outside broadcast truck no longer need to work out of that because they're working back from the remote location. So it allows you to also adopt a more efficient vehicle type there.

And so now stepping back, we're sort of in that micro view. So if we step back and look at a macro view of how these two workflows compare with each other, you see on the top, we've got our stadium, we've got our outside broadcast truck and then we've got our production crew and then it's common within the industry to have our production crew fly to site just due to production timelines and distances.

But now if we compare to live cloud production, we'll see what's similar and what's different. So similar. Same is the stadium, the cameras that stays the same. But what starts to change is you start to have the opportunity to shift to that smaller transport vehicle doesn't need to transport as much gear. Maybe you utilize a fly pack in order to ship the required video encoders that you still need on site. And there's still gonna be some people that travel the site, there's still gonna be some on site production crew in order to hook up those video encoders set up the video cameras, but where it starts to deviate is then you now are utilizing AWS to host those uh virtual instances in order to run the production.

And by running an AWS, not only are you gaining advantages of the scalability and flexibility that just come naturally from running an AWS? But you also gain some of the benefits I mentioned earlier, which is the carbon efficiency and the renewable energy benefits that we mentioned as well from that sustainability of the cloud aspect, but most impactfully and looping back to our point about Albert is this can allow an entire class of people to no longer have that requirement to travel entirely.

And with Albert saying that travel remains the largest part of a production's carbon footprint. This is really our target area of rethinking the way we're doing business such that now that group of people no longer needs to travel to site. And if we think about air travel, in particular, air travel has a high carbon intensity factor and happens to be one of the easier modes of travel to estimate which we'll go into here.

And so here as I wrap up my set part of the section today, well be this can inspire you a little bit about looking at the way that if we rethink our workloads and rethink the way we're doing our business. As usual, we can start to have a big impact. And so here's a very basic way of calculating estimating just your avoided emissions from air travel if you were to go about this workflow.

So what do you do? You take total number of people no longer traveling, you multiply that by a carbon intensity factor that sounds complicated. But there's lots of free calculators online that are trusted sources. And what you can do is input your source and your destination of where you're flying to. You get an a carbon intensity for that mode of travel, you take that amount you just calculated and then you multiply it by the total number of games that you're gonna repeat that for. And when you come out with is an estimated amount of avoided emissions that you've saved.

Thanks to switching to this workflow. Now just to put some, some placeholder numbers in here, these aren't the actual numbers from the European League of Football just but just to give you an idea of what that would look like. So you say you had five people like your technical director, your audio engineer, maybe they're no longer traveling and then you multiply that by a carbon intensity factor.

So here's an example of one from a flight within Europe and its 0.127 metric tons of CO2 for a round trip flight. And then multiply that by total number of games. Now, I will say for the carbon intensity factor, it's gonna change from destination, a longer destination is gonna have a bigger carbon intensity factor. But for the sake of this example, we'll keep it static, multiply it by the total number of games. And then what you come up with is 28.6 metric tons of CO2 avoided.

Now to put that number in context, the International Energy Agency has gone through an average doubt. What's the average carbon footprint from energy for people that live on earth, human beings that live on earth? Individuals like me, you and the number they came up with when they averaged out across the whole world was 4.7 metric tons of CO2 per year. So you can see from this example that's per year. And with this small change, you can start to see the size of impact that we can have.

And so I think if you apply this type of thinking towards your workflow and your different types of you apply it to live cloud production. But also you apply this type of thinking towards other parts of your business that may be carbon intensive. And you can start to see how our industry can start to do our part in order to stay within that carbon budget.

And so the realization of that estimated carbon footprint was exactly what the European League of Football discovered this year when they switched to live cloud production. And so to tell you more, I'm very excited to welcome Gelco, their CEO to the stage. We'll tell you more after we watch a quick one minute highlight from the league's 2023 season.

Get off and please give Joe Call a round of applause as well. Thank you all. Hey friends, one family out here, right. Fall out. We just come make a name for yourself. We just start, we just, and mm, understand what is going on here. Here he goes. Who that passes intercepted that is made he's taken down. We can go the outside, leave it. Oh, yes.

Thank you for having me here. Um normally I would, uh think you expect somebody standing up front here and talking about soccer and not about American football coming from Germany. Uh I'm sorry, but uh I want to talk a little bit of American football because uh we say it's the next big thing in Europe and especially in Germany, uh is the biggest market.

And uh uh what happened over the last 10 years, um is that uh American football, um came up from a niche sport to, to a major sport in Europe. So when I was working at TV station, the biggest TV station in Germany, we only showed the Super Bowl a year and everybody was complaining, why do we show this? Nobody's interested in American football anyhow, we don't understand the rules and why did we, did we show that soccer is much more interesting, although we play 90 minutes and at the end it's 00 and everybody's excited.

Uh this was something that uh no American ever understood. But uh uh that's how the world was running. But when I left PO, we showed over 150 hours of live coverage of the NFL in German TV. And now it's the second most important sports in Germany. Uh you may know that the NFL had two games this year in Frankfurt. And uh if we would have the capacity in the stadium, they would have sold more than 3 million tickets overall uh for these two games. So pack, it's, it's amazing what is happening overall.

And that was the reason why we said, ok, maybe. Um there is also some kind of headroom with American football to do their own league and not only from the amateur status uh view but also thinking about a more professional league. And that was the reason why we started with the European League of Football in the middle of COVID 2021.

Um and uh yes, started uh in 21 with eight teams from three nations. So it looked like a small European league uh but uh mainly driven by Germany in 22. Uh we expanded uh to uh to five countries and last year we finished the season with 16 teams from nine countries. So overall, I would say it's a real European league.

And uh as you see here, um we had uh a record breaking attendance with 32 32,500 people in Hamburg for one regular season game in American football in Germany. We played in the soccer stadium of, uh, the house F so things are becoming big, uh, but also, uh, what is becoming big as well is, um, that we have to think about, uh, what was mentioned about traveling more games.

We started at 43 games in the first season, had 70 games in the second season and last season, close to 100 games. So the things become more and more complicated and uh yeah, you have to think about production and the way of production, especially, um when you're, when you're getting bigger.

So, um when we just, uh just a short view of what, what we are doing. So, um when we, when we talk about emotions and, and fans in Europe, over 100 million people, uh are more or less now dedicated American football fans. It's not as big as in America, but uh we are, we are improving and it's becoming bigger and bigger as I said, second most TV sports uh at the moment.

So this is something we have to take care because people compare us uh with the NFL or college football. So also from the production standpoint that this is what we always should have in mind. Uh one big thing is, um, it's compared to soccer, it's a family sport. So, you know, maybe you saw the pictures from Germany, from other countries through Europe.

So you have a lot of things with on the one hand, fans on the other side, hooligans. Uh this is something that is, that is not, not a fact in American football. So families are coming in, they are celebrating, they have fun together and spending up to seven or eight hours during one game um in a, in a stadium.

And uh we we are diverse. We have players from 45 nations playing in our league and um we are doing do more and more life experience. So live games, as I said, um we, although we are young league, uh we are running um betting business. So data and statistics is really important for us when we are producing, we have our own TT platform running our own game pass that is available throughout the whole world.

So uh from the, from the, from the production standpoint, we have to be professional in every way and be sustainable in the future with all the production of our content. Um you see once again, um as we started uh small and uh but we are thinking about 25 maybe even 26. The, the goal that we have is to be present with 24 teams through at least 15 European nations.

And with that covering more or less uh half a billion of people uh with uh franchises in this country where we are, where we are. So this would mean as you see a lot of uh producing a lot of traveling around. So we will have at least at the end 160 maybe even 200 games at the end.

So, um it's all about production and uh traveling through Europe and uh this set, um our starting drawing from the second to the third season was ok. We are becoming bigger. We have to produce more. We have to be more on site, traveling traffic. It's summertime holidays, people are um traveling all over the world and all of a sudden, then we are also traveling with 1015 people uh next to the teams uh to the pitches.

So this was a point where we said, ok, we have maybe to uh to do a break and to rethink um if we want to produce the same way the other leagues produce over the last 50 years. And as we are young, uh I thought, ok, uh maybe it's a good idea. Um um as we think that we are innovative on the one hand and on the other hand, uh we are young.

So if we do some mistakes in the beginning, maybe it's not that hard uh like for a league that exists over years.

So, and this was the reason why we said, ok, how can we do this? Uh how we can, how can we do remote production and then who's uh technically capable of implementing this? And uh as I said, we were start up. So uh innovation means sometimes being fast and then rethinking and uh uh after planning of 34 months with the other things that we have in mind.

So we said, ok, we want to change it all of a sudden. So at the end, we had it eight weeks before the start of our leak and the June this year who's able to do this within eight weeks, preparation time.

Um yes. And also thinking from now is who's capable to do this, let's say also the next steps with us integrating data health um things into the feed into the production, so um be sustainable also the future and not only a lucky shot for one season and then coming back to the old way of production and for sure again, start up.

So we are not the NFL or even not the DFL from Germany, from soccer. So we have to take care about costs, we have to be cost effective. So that was also thinking, ok, who can help us? And this was at the end, the reason why we said, ok, we want to partner with AWS.

Um we want them to become our partner in order to start in 23 with this kind of remote production, but also be able to grow with them together in the future. This is uh a child that i love most. It looks so easy, it looks so sweet and uh but it means so much sweat uh over the season.

So i'm happy that i don't have to explain this, this will be Felix uh uh part. But one thing is uh we had uh we had this season sometimes four parallel games from four different countries. So we got 2830 signals at the same times into our headquarter in Munich.

And uh as Jason said, we were where we have to mix audio video graphics, uh put a live commentary, we sometimes show the games with four different languages as we are the European League and not everybody is speaking English.

Um so to mix it with uh with different languages and then distribute it the normal way of the different ways uh that a possible. So um because uh although we are talking about remote production, everything, um the one thing the customer outside doesn't care about remote or SNG or whatever he just uh is taking care about is the quality good enough. And is this what I'm receiving at home? That what I'm paying for?

So, and as i said, this is my Swedish chart and uh um i think uh Felix will explain later how it worked, but one thing is uh we made it happen this year and uh this is uh uh what i was at the end proud of uh that we did it.

So um and the reason why i did is uh we had access to a technology that a start up like us never would have on its own. So this is, uh, we, we, we couldn't allow us but, uh, you working with AWS together, uh, make this happen.

So, and, uh, and what was really important that at the very beginning is i jump over to the learnings is, uh, we only had eight weeks, as i said, but, uh, you have to have a dedicated team that is working on that because it's totally different than producing the normal way because uh i would say uh if you have a lack of time that we had, then communication is all.

And at the same time, the communication is the biggest issue that you have because uh if you're doing something remote, but you're not that stuck in, in the live production. So if things delay for five minutes, it's not a problem. But if you have a tv, partner who is expecting you uh on the f to be on frame uh live on this tv screen and you have to say, oh, i have to delete for or delay for five minutes. It's a big, big issue.

So um communicating on the pitches on four different pitches in four different countries, maybe sometimes in different languages because that's, it's not, it's not us, it's so sometimes we have Austrians and you may think they speak German, but it's not always the way uh it's kind of dialect of German and sometimes more difficult than speaking Spanish.

And um so um you have to talk in different uh in different languages with the people um on the pitch and it must be a real time. So, because you put in the graphics and uh data and everything in real time, so you must know exactly within frames what is happening next because you put in a graphic and imagine you think it's a total, but all of a sudden it's a close up. So it works like amateurs uh working on a pitch if you don't, uh if you're not able to um to put it together.

So at the end, we made close to 70 games from 99 this season by remote. And uh i would even say, although it was an adventure, but at the end, uh we even improved the quality of that what we have shown to the people, especially through our game pad because we put in um more data statistics to defeat.

And we also had a multi camera view that was allowed. So you was, you were not only able to watch a game on one mixed feed, you were also able to watch maybe from four different camera perspectives at the same time and simultaneously.

Um what was a great reason for the people to pay the €100 for the season pass overall. And at the end, uh we said here, we saved more than 40 tons of carbon um this season.

So, um i would say, and this was the reason also uh to work together. Um this is not the end, just the beginning. So as i said, we want, we want to become bigger. We want to play in 15 countries through Europe. We want to play 160 to 200 games a year.

So uh we want to be more innovative than other league in Europe overall. So we have to optimize a lot of things and uh implementing the learnings that we had over this season.

Um all matches must be produced. Also the top matches that we called from tv site this year. Uh we didn't produce by remote because the tv station said we, we don't want it. It's, it's, it's too difficult and we don't believe this will work out.

So now we have to build trust with this kind of production and also that the big tv channel say, ok, this is a way of production that we are accepting and that you are allowed also to deliver us the signals via remote production data, statistics, betting, everything is becoming more and more important.

So again, putting into feeds real time and implementing this into the feeds is 11 big thing. And uh we see here, second screen application, our app and everything, you know, this stuff uh by the, let's say normal production. So what we, that's what we're doing also, but we have to put everything together in this kind of environment of pro production and again, also one thing is american football.

It's the same in Europe, like here in America, it's uh athlete health. How can we use the sensors and everything with the with data feeds in order to improve? Uh um the athlete health is one topic that we are facing.

So and last but not least from my side. Um one big thing is, and this is something that we worked on after we made the pictures happen. So is from our point, um we want to earn more money if we wanna be more effective. So we need more data driven business and this is where the profit will lie in the future.

So thinking about remote production, working together with the cloud system, um the things with data, how to put everything together, how to implement this and at the end, how to capitalize is one big thing, not only for us but also for the other leaks.

Um i'm 100% sure that although we are maybe at the moment the first one, but the most uh small leaks niche leaks will follow because i think it's the only way to produce in future, cost effective and uh to reduce the carbon footprint that if people think serious about the thing.

And uh and in the end, i think it's also on us because the quality is even improving. And ii i really think even for the people that are traveling normally uh to the pitches uh it's more convenient uh to stay in Munich and to work from there than traveling in 24 hours to Barcelona stack in traffic and coming back producing carbon.

So i think overall it's, it's a benefit also for them. Yes, it's from my side. Thank you very much for listening. If it was a short overview over over the lelf. Um yeah, maybe next time, a little bit more. Thank you.

And after good morning, welcome at we. So my name is Felix Wicker. I'm a solution architect at Amazon Web Services. And during the last year, i accompanied the European League of Football when they were building their life crowd production pipeline.

So I'm very interested and I'm very excited to join today's session because it actually takes us on a very special journey. So it started with chasing telling us about the essentials how we can actually use technology to build a more sustainable business. He explained us how we can actually use the concepts of sustainability of in and through the cloud and therefore gave us this feeling like, oh, now i start understanding and i start seeing the opportunity for sustainability.

And then the journey went ahead with chalco coming on stage with music, bringing in some emotions, giving us some action and actually showing us in a very impressive real world example. What's possible if we think big for sustainability.

And with that, we actually arrive here right now right here because now we have the feeling. Yeah, i want to go for it. I want to do this. So i have to understand how it's actually built. And that's exactly the third part of the session telling us, hey, how does the architect, architecture look like? And what are the building blocks? I can lie on to get us started.

I actually brought you a high level timeline of the project. On on the very left side of the slide, you can see a starting point. So the starting point was mid of April when we had our on site visit at headquarters and were initially communicated about the idea of a live cloud production platform.

And if you look at the right side of the slide, you see the end point which is the first weekend of June because there, there's the first day, there's the first match day of the 2023 season. And now we immediately do the math and you notice from the start point to the end point, it's 7 to 8 weeks, which is an incredible ambitious timeline.

And on top of that, you have to see the conditions of this project. This is something we will from scratch new. It's something we haven't done like 2000 times before. So we can not really rely on routines and on top of that, it's highly visible. So people are going to watch us. So we definitely want to be successful with that.

So overall if you put all these conditions in a nutshell, we have to understand how we can actually derisk this project while accelerating it. And that's a super important thing because it's highly influenced by the way we build our architecture.

So let's look into it. What you see on the slide is a high level overview of the architecture of the elf slide cloud production pipeline. It starts on the very left with the signal production. So that really happens on site, you get to, you go to the stadium, you take your cameras and you produce your signal.

And then the next step is to think about a way how you can actually transfer that signal from the stadium into the cloud. And once you've done that you become operational because then you signal, you do the signal processing, you add the audio, you do your feed distribution.

And of course, while all doing that, you also build a virtual control room and you take care of the data management. So what we're actually going to do on the next slide is taking all of these elements and look what's remarkable and what patterns we can use to be successful like e was.

So let's start with the signal production. What you actually have to understand to really cross the point of signal production is what stadiums we're using here. So what we are not using is a stadium that you see when we build or when we watch the super bowl. So this is not a 100 k people stadium state of the art top notch infrastructure instead we have solid stadiums, but we have to think about ways how we get our signal from the stadium to the cloud.

So that's the key question here. And the solution for us to do that is actually to rely on 5G transmission. So we rely on mobile networking and we actually do that with 2800 units from LiveU. So why is that important because it allows us to decouple signal production, signal transmission from the stadium

So we are limited eliminating our limiting aspect of the infrastructure on site and we are becoming operational in the cloud. So 5G for us means a gateway which is flexible and adaptable and makes us operational. And with that, we move into the cloud which allows us to use the state of the art technology, the scale and also the sustainability of our operations.

Talking about sustainability, Jason introduced the concepts of of in and through the cloud. What we look here is often in the cloud. So sustainability of the cloud for ELF means that we use dedicated instance types. So for example, for signal ingestion, we have instant types with highly efficient network throughput. While for example, for signal processing, we use instant types which are built for graphic intense workload. And with that, we have the chance to run those workloads highly efficient and we are not only running them workload specific with high efficiency, we also run them in the Frankfurt region for the year. And the Frankfurt region is also one of the regions. Jason mentions which power consumption is actually a reputable to 100% renewable energy from there.

Moving on to sustainability in the cloud. And if we want to have understand the sustainability in the cloud, we have to look at the schedule because like Jason and Chao mentioned football actually happens on the weekend. So how great would it be if you whatever you build here, you can turn it off on Monday, Tuesday, Wednesday, Thursday because you don't need it. There's nothing happening and ramp it up for the weekend. So that's exactly what ELF does they ramp it up on, let's say Friday evening, get ready for the weekend, produce Saturday and Sunday and ramp it down on Sunday evening, Monday morning after the post processing. And with that, it's not only about computation, but we can also go one step further, which is about storage. So for example, in the case of ELF, we use three life cycle policies which allows us to refer to very efficient storage policies.

Staying a little bit with those building blocks around signal and cheston and signal processing. Very remarkable aspect of this architecture is actually third party integrations. You see a bunch of the most important partners. And let me be, let me be crystal clear with that, with our partners and without the ecosystem route AWS this ambitious timeline and this highly complex project would never have been realizable. So an incredible job of our partners. Thank you very much. I saw some faces here. So thank you. Yeah, that's great.

Besides that, there are also two main reasons why we actually use 33rd party integrations. And the first one is that it gives you a lot of speed. So instead of building everything by yourself manually from scratch, you can take off the shelf solutions from your partners and integrate them seamlessly. And with that, you gain a lot of speed, which for example allows you to build a virtual control room within seven weeks. So this is exactly what Jason referred to. This is the virtual control room of ELF close to Munich. And it shows you exactly what happens when you virtualize your hardware. So for example, here you see to the audio mixer during a match day.

The second aspect why we use third party integrations is actually because it de risks your project first one on the technical aspect. So by referring well-established industry solutions, you gain confidence you gain routine. So those are well established, well trained, well rated. The second aspect is actually maybe even more important because it's the human aspect, live cloud production and the introduction of live cloud production is a kind of transformation for your operations. It's not as easy. So you have to take care of the human node about it. And that's a great thing about third party integrations because let's think about the operators. They were maybe running with a T previously. They don't even notice if they do it remote or in the cloud. They only think they notice, hey, this is actually the same program. I'm used to the same tool i'm used to work with. So i'm actually not afraid of this transition. I couldn't rely on genes. I can work as I'm used to it. So it gives me confidence and a lab and perform at the best level.

With that being said, we're actually not too bad. We are already close to the goal. So we have speed, we have reliability. The only thing we are missing is actually scale. And you heard about childhood, this league is growing immensely and it's growing very fast. So in the 2023 season, we produced more than 60% of our games remotely. So if you would have to do all of that manually hand by hand, that would be a lot of work and a lot of complexity. So the solution to cope with that increasing operational complexity is automation. And that's exactly what we did for the feed distribution.

So we built the serverless start stop mechanism for feed distribution based on AWS Elemental Media Services. Looking into that we actually have the one for the starting defeats what we do is we rely on AWS Elemental Media, Connect to orchestrate our operations. We then use AWS EventBridge to detect events and to transfer them into actions. We use AWS Lambda to execute those actions. And then we add AWS Elemental Media Live to produce the signal, for example, RMP. And we actually follow the same mechanism, the same pattern when stopping the feet. So again, we use AWS Elemental Media, Connect use EventBridge to build a decouple and advent driven architecture, you Lambda Media Life. And also we are adding Media, Convert out of the AWS media services to build a flexibility and variety and formats.

And we're diving deep into this because it's actually super important aspect. Why? Because those are managed AWS services. So your producer, your producer, your production company does not have to become an IT company. They can still focus on what they are really good at because they don't have to worry too much about the IT underneath. So instead of giving them a lot of undifferentiated heavy lifting, you let them focus on what they are really good is good at producing highly fan engaging experiences.

And with that, we actually went through the whole architecture from the signal production at the stadium to the automated feed distribution. And we saw how we can build such an architecture in a highly ambitious timeline, quite a lot of learning. So let us look at the key takeaways.

First of all, Chao already mentioned, it achieving life, cloud production is a huge success, success and gives a lot of opportunities. And that's the point. It's not the end point. It's the starting point. It opens up an incredible big landscape of opportunities because you now have accessibility to technology you probably never seen before.

The second thing with all those opportunities, you have to take decisions and you're moving fast. So you have to take decisions fast in order to achieve the best outcome on all decisions. You have one guiding principle and that guiding principle is to be fan centric. So whenever you have to take a decision, ask yourself what's best for the fan, what's best for the broadcaster, what's best for the customer and then work backwards from that, that highly increases the chances of taking the right decisions and last but not least.

And this is important for the session today that we have to do a shift in mindset. So historically, we might be thinking of economics, performance and sustainability as a trade off, right? If I want to be performed, it might cost me a lot. If I want to be sustainable, I might have a decrease in performance. We have to change that mindset because on the case of ELF, we actually see that you can have all of that and you can have all of that at the same time as they are complementing each other. The only thing we have to do is to use technology wisely.

And with that learning and with that change of mindset, we get some extra motivation and talking about motivation. We arrive at the most important slide of today, which is the call to action. So Jason Shaco and me actually want you to do three things.

The first one is realize easy wins by leveraging the sustainability of and in the cloud. Those are concepts introduced by Jason and you can realize those easy wins by talking to your account manager and saying, hey, the account manager, i want to have a well architected r with a focus on sustainability. That's a super direct way and a super accessible way to make your first impact. So i will ask is to schedule a well architected review in q 1 2024 with your AWS account manager focusing on sustainability. If you've done that, you actually had a great start because you improved the sustainability of your it and your it landscape.

But we want you to go further. We want you to use the sustainability through the cloud to be bold enough to re invent the sustainability of your business. So have a really nice week at weekend at rear rent. And after that, you go back home, take your road map for 2024 and ask yourself in this road map for 2024. Where are we actually using technology to build a more sustainable business we want is not an acceptable answer. But i think with your motivation and all your learnings and all your inspiration, we will find plenty of opportunities to transfer this knowledge into the road map for 2024.

And the last thing is when you're going to do that, remember you're not alone. So this is about sustainability. We are in this together and we're going to solve it together. So let's start solving it together right now.

Thank you very much. We brought you some extra resources so you can access them via the q articles, for example, the well architected framework on the sustainability pillar. And also we appreciate very much your feedback to via the app.

Thank you very much.