Klaus Koefoed spent more than 25 years in IT services and consultancy, at Capgemini, Deloitte, and Accenture, before taking over as CEO of T-Systems in Northern Europe. In this episode of the Smooth Scaling Podcast, Klaus walks host Jose Quaresma through the rise of sovereign cloud: why a theme almost nobody raised two years ago turned urgent in early 2025, and what actually changed. They get into Europe's real position against the American hyperscalers, why the answer is rarely either/or, and how leaders should weigh the data, operational, technology, and legal layers of sovereignty. Klaus is candid that none of this comes for free. Multi-cloud adds complexity, and the right answer depends entirely on who you are. The back half turns to sovereign AI, where T-Systems and NVIDIA have stood up a billion-euro Industrial AI Cloud, and where Europe's gigafactory ambitions, real use cases like simulated wind tunnels, and AI Scrum Teams come in. A grounded, practical look at building and running infrastructure when geopolitics is suddenly part of the architecture.
Klaus Koefoed is CEO of T-Systems Northern Europe, leading Deutsche Telekom's B2B IT-service business across the Nordics, UK and Ireland. He joined T-Systems in June 2023 from Capgemini, where he was VP and Nordic Head of Cloud Strategy & Transformation Advisory. Before that, he was a Partner at Deloitte Consulting, with earlier years at Accenture. Copenhagen-based, with 25+ years in IT services and consultancy. His move from advising to operating arrived just as digital sovereignty stopped being a compliance footnote and became a board-level resilience question. Under Klaus, T-Systems Northern Europe has pushed hard on Sovereign Solutions for Europe and especially T Cloud Public (formerly known as Open Telekom Cloud) — now ten years old enterprise grade European Public Cloud — and recently opened the Industrial AI Cloud, Europe's largest sovereign AI-infrastructure, with 10,000 GPUs connected to the existing portfolio. That makes him one of very few executives in the region who can credibly talk about running sovereign cloud and sovereign AI at scale in Europe.
Episode Transcript
Jose:
Hello and welcome to the Smooth Scaling Podcast, where we speak with industry experts to uncover how to design, build, and run scalable and resilient systems. I'm your host, José Quaresma, and today we had the pleasure of chatting to Klaus Koefoed, who's the CEO of T-Systems in Northern Europe. We had a great conversation with Klaus about sovereign cloud, the current state of the cloud industry in Europe, what is still holding it back, and how far it has progressed in the last few years. We also dived into the new challenges that AI and LLMs bring to infrastructure choices, and how T-Systems with the Open Telekom Cloud is investing in that area. If you like this episode, please subscribe and leave a review. It really helps the podcast. Enjoy. Welcome, Klaus. It's great to have you on the podcast.
Klaus:
Thank you, Jose.
Jose:
And I would like us to start straight in, and maybe a little bit with your background. So you spent more than 20 years in consulting, at Capgemini and Deloitte and Accenture. And we'd like to hear a little bit about what made you make the move from the consulting world to the operational side with T-Systems.
Klaus:
That's a good question. I think I saw a very interesting opportunity outside, you could say, our region. We work in the Nordics, or I work in the Nordics and UK. But in the home market, we're sort of the big animal on the savannah. We're not here, but the solutions we have are extremely relevant also for our market in Denmark and abroad. So I thought this journey of making that much more visible to the market, and being something else than what we might have been known for in the past, was a cool challenge.
Jose:
At the time when you made that move, was there a sovereignty theme? Were you hearing that from client conversations?
Klaus:
No, not really. It was more of the usual topics around security, of course. They are part of almost any conversation. But sovereignty as a theme, no, not really.
Jose:
So when did that explode?
Klaus:
Yeah, it exploded around January '25, I would say. For us, it started really with the Ukraine war, when the access to satellites was turned off for a while for Ukraine. We thought and discussed internally, so what if you sort of extrapolate that scenario onto other technologies like public cloud, like other themes? What will the impact then be? And then we started talking to clients about that risk scenario, really, which it is, right? What if that... And in the beginning, I was like, hmm, not so much. I mean, people had noticed, but that was about it. Then the International Criminal Court happened, where some of the judges were, and still are, treated like sanctioned by the US. So they didn't have access to Office and other things. And then it started gaining momentum from there.
Jose:
So the risk has been there for a while, but I guess it was made more real in the last couple of years.
Klaus:
I think the examples start adding up, and then the need to maybe have another round at looking at the what-if scenarios. What if whatever crazy scenario you can think of, if that happens, then what do we do? How do we ensure we can run our business? And then you can say, of course, the more off-the-wall the scenario is, maybe the likelihood of it happening is lower. Whereas other scenarios will be much more realistic, probably.
Jose:
And I think you've been thinking a lot, and writing as well, into how does Europe stand in this setup, right? So can you share a little bit of what you see are the main strengths, but also challenges, within cloud and resilience?
Klaus:
So I think, first of all, we as Europe have put our money elsewhere than in Europe for a long, long time. And I think it has made sense. You have some hyperscalers out there who are bringing value every day to their clients. Now we just have an additional dimension to focus on. So, as we've talked about before, for us it's not an either/or, it's both/and. Using what you already use for the right purposes, and then thinking this risk element into what you put where and why. And for Europe, I think it's an opportunity. It's an opportunity to maybe get a better balance of where you make your investments and put some of those investments into Europe. So we have a strong tech tradition anyway, and good innovative minds. So of course we can do it. Question is, of course, where do we put the money then?
Jose:
And what would you see are the top three challenges that Europe is facing right now?
Klaus:
So I think there's sort of a private and public sector challenge. One is, you know, how big is the problem really? And I think that's a healthy discussion to have, because if you are to invest in something, the outcome or the risk mitigation has to make sense, right? That you put your money where you can see there's a relevant return. And then in public sector, there's also an element of making sure that the requirements you put up are not necessarily only the usual requirements, which will then, and that goes for private sector also, that is not pre-tailored for the known past outcomes, if you will, but that you think in a slightly different way. And there's a lot of push, I think, if you read the press, to put in criteria around scoring mechanisms for when something is European. And it's, of course, finding a good balance. So I think for private sector, it's also a matter of, if you think this is important, well, then you also have to act in that way, right? So you have to make your selection criteria to at least address the theme of European platforms, as an example.
Jose:
What I hear is, you're advising everyone to take a pretty balanced approach. As you said, it's not either/or, but to have a thoughtful engagement with those ideas and make sure that you understand what the consequences are of doing it or not doing it, and taking it from there, right?
Klaus:
Absolutely. I think, you know, let's not be lemmings and just run in a certain direction because everybody else does it. I would advise anyone to maybe take a step back. I think there's some learning from the first waves of cloud, where everybody were excited. There were good incentives, et cetera. So maybe in hindsight, too much was sort of shoveled onto cloud. And then, of course, you have other companies like yourself who've built cloud native, which is different. And there are some learnings out of that. Now it's about taking a step back and saying, "Okay, what is my real need across my business and my application portfolio?" We as a company, T-Systems and Deutsche Telekom, we have so much legacy and so much investment in technology, we probably have it, right? So we are on the exact same journey, moving from old to new, with a different lens on what we place where. So we have a mix of American hyperscalers, our own European cloud platform, private cloud, and other platforms. So shifting the balance between those different platforms.
Jose:
What would you say is the ideal scenario for a company that is a SaaS company out there right now? Do I hear you saying it's more about being able to play, or to be in whatever cloud makes sense? Is that a fair assumption? Or how would you frame the perfect situation right now for a SaaS company?
Klaus:
So now you get the ex-consultant answer, right? The answer depends on the company you're seeking the answer for. Because if you're a big telco like us, it might be manageable to actually be on several clouds and private cloud, etc. For a cloud native company, more of a startup, you really have to think hard about, "I am where I am for a good reason. Why do I want to dip my toe in platform number two or three?" That must be because either the market simply demands it, or the risk profile of, there's a lot of talk about kill switches, right? If that happens, then what does that mean to my business? So you have to look at, I would say, per company what the right answer is. And we all know that every time you add a new technology, your people have to learn it. And you have to manage that additional complexity. So it's about finding the right balance, again, depending on who you are.
Jose:
So let's say there's a startup just starting to develop their system right now, right? It feels like it's quite an interesting thing to think about, and a discussion to have, because of course you don't want to overcomplicate from the beginning, because you're still probably trying for market fit and trying to understand if what you're building makes sense. But on the other hand, if you would start with at least a multi-cloud perspective in mind, then it would be easier at some point later if you want to have that option. So have you seen that balance out there recently?
Klaus:
I think the concept of optionality is really something you need to think about from the beginning. Because if you go all in cloud native, obviously that gives you loads of benefits, but it comes at a price, right? Because that gives you less freedom of movement. And if freedom of movement is important for whatever reason, then maybe you have to think a little bit differently.
Jose:
And would you say, can you think of any case right now where freedom of movement is not important? Because with this current context, I feel like it would always be important.
Klaus:
It is important. And we've seen smaller SaaS companies that went hyperscaler-native from the beginning. That was obviously a quality sign when they were out in the market, and still is, of course. But then you go to certain markets where actually that's American, and we're not sure we like that. But if you come back when you're on something European, then we'll talk again. And that has changed quite quickly, I would say. And that, of course, means that you have to think about how do we not just make our life incredibly more complex by being on two or more different platforms, but think your architecture through in a good way.
Jose:
That makes sense. And you mentioned previously thinking about these what-if scenarios, right? In your experience, I guess it's probably the CIO usually, when they first think about these what-if scenarios, is there anything that has surprised them when they start thinking about, okay, what if we lose access to this hyperscaler? Has there been any kind of finding or thing that has been unexpected, do you think?
Klaus:
So for us, we have those kinds of conversations. I would say, if you want to do it properly, you either need to do it with your own risk team, or find an external advisor to help you work that through properly. I would say often you run into, "Well, will that ever happen?" No, no, but let's park that. Because now we're just walking through what could some of the scenarios be, and it's about resilience at the end of the day. How do you make sure that you can run your business even if you don't have access for days or longer, right?
Jose:
And so, yeah, it's that aspect of, there's the likelihood or the probability of it happening. That's one aspect of it, and those have, I think, in some of these what-if scenarios been going up and up. But the impact of it happening, that's...
Klaus:
Exactly. And that's normal risk analysis, right? And some of those off-the-wall scenarios before '25 might not be as crazy as we thought back then. And then, of course, the impact can be high: lack of access to your core business systems, or data, or IP, et cetera, et cetera.
Jose:
And when talking about sovereignty, you usually talk about it at three levels, right? It's data, when we get to the straight data, operational, and technology, right?
Klaus:
And then you can say legal framework on top of that, right?
Jose:
Can you tell us how should leaders think about those three levels and figure out which one do they need?
Klaus:
Well, I think at some level you have to think about all of them. And then I would say there's an element of a journey, in the sense that if you take our public cloud, or any hardware out there today, it's either Asian or American. We are behind in Europe really building. We'll get there, but right in this moment of time, it's sort of either one or the other. And then, of course, on top of that, there is the rest of the stack. Is it open source? Is it something else? And where's your data sitting? Is it in Europe? Does it stay in Europe? When you have sort of peak hours on a cloud, does it suddenly shift to do stuff outside Europe or not? That's quite important to know. Then, of course, who is it that operates your platform? Does that again stay within Europe or not? And then you would say the legal framework, or things like the CLOUD Act. If you somewhere end up being an American company, then that applies, full stop. It does not apply to us as a European company. So no one can force us to give anybody your data. And then again, as a journey, you can pick solutions today that will give you functionality that is very close to what you see from the American hyperscaler. So you can get very far, very quickly, very easily, actually, in terms of moving from A to B. Have the market, broadly speaking, in Europe solved that it runs on a European box? No, not yet. But it's a strong step forward already now that that's available.
Jose:
And maybe just before we go into some more of the AI side of it that I would like to touch upon as well, I'd like to talk about, let's say that someone is starting to look at this seriously and they actually want to look into the options that are available in Europe, right? What would you suggest for the main evaluation criteria? So what should people be thinking about when looking at the possibilities out there?
Klaus:
So we have spent quite a bit of time since January, February last year, talking to clients about — we have one slide showing where you have two dimensions. You have functionality on the y-axis, and then you have sovereignty on the x-axis, right? We as a company, we serve clients from anything hyperscaler to sort of bare metal servers, right? And in that continuum, you have to find out with yourself. So if I have something where I want to leverage AWS full speed, that's fine. But over here, it may be so secret that I'll give up any kind of functionality. I'll come with that more or less myself. I just need a super secure platform, right? So that's the starting point. What is the need you have? Either very granular, or as a more holistic view. And then there are several European options out there. So it's really, are you looking more for sort of, you could say, sovereign servers, if you want to call it that, or do you want something that is much more like what you're used to? And then there are different players across these different solutions.
Jose:
But it does start from knowing and specifying what you need and what you want, right? You have to understand your own need and your own situation first.
Klaus:
That's the best place to start, because otherwise, how do you know what you're looking for? And then I would say there's a lot of noise out there. There are loud voices. There's a lot that is being labeled cloud.
Jose:
Yeah.
Klaus:
And then, of course, we've seen several clients do that, you know, they start with maybe 15, 18 European clouds. And then they have defined their own criteria for what they're looking for and their needs, and then they work their way through, and then they end up with two, three answers, and then they do proofs of concept on those, and then decide based on that. And not everybody can do that. But I think my point is, at macro level, you can call something cloud. But then if you start looking at, either bottom-up, what are the services available, what is the security that is there, what are the managed services, service levels, security certifications, etc., then you suddenly begin to see some of the differences of the possible answers out there. So it's not less complex, unfortunately, for clients.
Jose:
No, no. But that's, again, back to, it's not coming for free. And part of it is also the complexity, and of course, discussing and setting something up to be able to run in several clouds will inevitably have to be more complex, right? Well, I don't know if "inevitably" is the right word, but most likely will be more complex, right?
Klaus:
Yeah. And then, of course, if you're a sizable organization, typically you will have an orchestration layer saying, you know, based on these criteria, this thing lands over here, or we're building services for clients over here based on AWS, or based on our platform, or based on a third thing.
Jose:
Yeah. And I think, to be fair, some of the bigger companies out there have been doing that anyway at some level, right? Maybe before, everything was in their data centers, and they started to look at cloud and the public cloud providers. Now it's kind of continuing that spectrum of possibilities, just adding some of the more nuanced discussions around Europe or not Europe.
Klaus:
Yeah. And I think, like you, the recommendation is: have a look at your insurance at home every year, right? Go through, am I insured the right way? I think you need to do that also as an organization around cloud and platforms in totality, to say, given we are where we are today and the world as it looks right now, do we need to fine-tune? You probably need to look at it more often than that, but at least one year. These days, one year is very long. But it's more of having the discipline to say we need to look at this on a recurring basis, and not something that is triggered from outside events, but that you have it as a discipline.
Jose:
Yeah. And you mentioned that some of the organizations have this orchestration layer that helps them navigate, I guess, simply put, where to put the applications. So are you thinking about a concrete layer, a piece of software or a system, or is it more of just a decision diagram, or a bit of both?
Klaus:
I would say it's a bit of both, right? Of course, it needs to be technology-supported, but that you have predefined, you know, okay, if it's this type of data, for instance, then the right home for that is over here. Or if we need to build the latest cool thing facing customers, then we typically leverage this hyperscaler for that, for these reasons that you have thought through, and then it's orchestrated sort of in the background for the business.
Jose:
That makes sense. And also, back in my consulting years, I did work quite a bit with Red Hat and with OpenShift, and I've lost a bit of track of the work that they've been doing the last couple of years, but I think a lot of it was that orchestration layer that was already being discussed at that level, right? That, well, you develop something, and then you can decide where do you want to put it, right?
Klaus:
And one thing is, of course, geopolitics and all that, but there's also an element of freedom of movement. If you have that, then if you're over here today and prices evolve in a way you don't like, well, then at least if you have set yourself up to have that capability to move, you can decide to move, or you can use that as a negotiation factor.
Jose:
You mentioned the insurance, you used insurance as a metaphor, and here I also feel like it's a bit the same paradigm, right? So you might be paying the cost upfront of having that possibility, but then you might use it or not, right? So I guess that parallel is also there. And we talked about hybrid cloud, but do you see any difference between how your customers use hybrid cloud and multi-cloud? Does it make a difference in the work that you do? Or how do you see the difference between the two?
Klaus:
Yeah, good question. Actually, I'm not sure. I think there has been, especially in our region where we're sitting in Copenhagen today, in Denmark, there's been a lot of appetite for public cloud, right? In the Nordics in general, I would say. I think now we sort of see a wave two. You could say the learnings are there from X number of years on hyperscalers. And commercially, also, how much do I actually leverage the platform services versus more using it in some areas as infrastructure as a service? And I think that also helps fuel the considerations about, so in my next wave, what is then the right answer for me? I don't know if that really answers your question.
Jose:
But yeah, it does, it does. Thanks. I think that's, again, back to thinking about the risks and what you want to achieve, and having an overall tough discussion and tough work on considerations about that, right? And figuring out where to go from there. Maybe it sounds simple when I say it like that, but it is, I think, very hard to do.
Klaus:
But on the other hand, if you think it through well with what you know today and then execute based on that, and then of course it shouldn't be that today we go left and in three months we go right. You probably need to think a little further than that, so you don't drive complexity and cost through the roof. But I think in this wave two or three, whatever it is for organizations, there's a time to reflect and take some of those learnings of being on cloud and say, okay. I mean, for us, again, T-Systems and Deutsche Telekom, we've been on a journey where we put things on public cloud, and in some areas you leverage it to the fullest extent. You learn that, okay, but actually this application is just humming away quietly, and we're not really leveraging the platform services. So why is it running there? Is it commercially viable to just stay? Or would it make more sense to move it to a private cloud platform? And had you said that, I think, two years ago, I'd probably be called old school, but actually we see in several clients that putting your pieces of the puzzle together in a slightly different way going forward, and private cloud, seems to be a hot thing as well.
Jose:
Yeah. A lot of our discussion has been around the risks, but there's also some cost perspective as well that I think has been a bit more discussed in the last years, of actually going a bit back from public cloud into the private cloud. Again, depending on how your application is used, right? If it's, for example, a lot of very variable workloads, then I think public probably has big advantages there. But as you said, if it's humming along there, then there's probably something to save for having it in private cloud, right?
Klaus:
Yeah, yeah. Maybe, yeah. There could be. And also, as I said, if you really need all the services that are there, then public cloud is the right answer. Whereas if you don't really, then maybe it's not. I think that's exciting, you know, all these outside impacts that come, plus running your own business and looking after it. What is the cost of having this thing running to serve our clients? And can we optimize some things across that whole ecosystem? And of course you can.
Jose:
And I would actually like to jump a little bit over to AI and sovereign AI. I think we couldn't have an episode these days without talking about AI, right? So can you share with us, how do you see the European alternatives to AI? How do you see them measuring up with the American ones, and where do you see the main gaps?
Klaus:
So I think the AI initiatives are obviously today built on American technology. That's just where we are today, right? So again, that will be a journey in terms of starting to produce something European on the hardware side. On the other hand, you could say as Europeans, we're quite good at being innovative. And, I mean, Mistral is one example of massive investment in Europe, right? Which I think is excellent. We are on it too. We have AI on our public cloud already. And we decided in the fall, together with NVIDIA, that Europe needed a sovereign answer as well. The EU is fostering this idea of five mega AI clouds, you could say. And that runs a formal process. We decided to invest because the need is there in the market. So first phase opened in February, right? And we have clients on board. And I think this whole movement of European options is a great one. That, back to your question before about sovereignty levels, gives you the same benefits as we discussed then. And we see sovereign LLMs being built, being available today, sovereign Gen AI solutions available today. And a lot of investment going into this region to build data centers, both for private, public cloud, AI.
Jose:
And where do you see — you mentioned you have customers running on your sovereign AI. Can you share with us what are the main, any specific use cases that you see being run? Or is it kind of justifying that?
Klaus:
Yeah, yeah. So that's evolving fast. What we have done is, you know, one thing is that you can buy a GPU per the hour, and some clients need that. But we have, like you say, almost a store on top, in the sense that we have partners putting their solutions on top. That means instead of talking about GPU per hour, now you have a business use case that you can then leverage these solutions to solve. We have a client, as an example, called PhysicsX, and they come out of doing wind tunnel testing for automotive, but they can do simulated wind tunnel testing for anyone, right? And optimizations of things you produce, for it to be more efficient, cheaper to produce, etc. So solutions like that, from them, from Siemens, from SAP, etc., are now put on the platform. And that means that you can then run whatever service you need, but in a sovereign setup. So you can bring your own data and feel comfortable that it stays here. And it's not just about GPU per hour.
Jose:
Very interesting. So if I am a customer that needs to simulate some wind tunnel results, then I can buy that service in your cloud, in a sovereign way, and run it there?
Klaus:
Yeah. So we see a lot about digital twins also. Digital twins on steroids, so multiverse kind of, you know, full factories etc. running SAP BTP. So you would say SAP AI in a sovereign setup. And some of the German universities have pulled up — there's an initiative called Soofi, and that is to build, like, the European answer, if you will, to ChatGPT, but in a sovereign way. So significant, you would say, parameters set underneath that, and that is progressing well as well.
Jose:
So what's next for — I mean, it could be for sovereign AI, but what do you see being the next step, or the next challenge being overcome there?
Klaus:
So I think there are two, three challenges. One is, of course, for us as a company, we want to bring more solutions like that from partners onto the platform, because that's where you move from buying sort of infrastructure to buying how can we help you with a business outcome. And I think that's a general theme with AI. We all know it's cool, but how do we move beyond the cool to something that makes a difference for society and for our customers? So that's one. And then, of course, that the market continues to move in a more European direction, in the sense that when public sector or private sector need to buy new, that they think about not only buying what they used to buy. And this is not a cry for, please come buy some of ours. It might sound like that, but I think if we really want maybe a better balance now between buying European versus buying what we used to buy, then the time is now, you know, to put your money where your mouth is, so to speak.
Jose:
And maybe getting closer to wrapping it up, before we get to a couple of quickfire questions, we'd like to ask, if we have a CIO or CTO listening and they're maybe starting to think about this more seriously, what would you say is the next thing that they should do?
Klaus:
If they're thinking more about European alternatives, the obvious answer is come talk to us, of course, right? Or at least spend a bit of time looking out there. What are some of the players out there? And invite them in for a conversation. I would promise anyone that invites in some of the bigger names from Europe that they will be positively surprised about the maturity of the platforms.
Jose:
And — thank you for the advice — is there one thing that comes to mind that they should stop doing, or not do? It's probably a bit of a harder question.
Klaus:
Yeah, what should you stop doing. I don't know if there's something in particular they should stop doing. I mean, I understand the complexity of even evaluating and potentially inviting somebody new in. That drives requirements in terms of what people are trained on and all that. But I think you have to sort of educate yourself about the possibilities out there. And then, with the themes we talked about before about risk and commercials, etc., find out what is the right answer and then what way to go. So I don't know if that's anything about stop doing, but maybe lifting the gaze and thinking European also, not instead of, but also, so that you are more inclusive in how you look at the option space for you.
Jose:
And I feel like also at least stop ignoring it. I was thinking about that, like at least they shouldn't just cover their eyes or continue as if it's such a small risk that it's not worth considering, right? It feels like it's getting to a point where you need to start considering things if you haven't done so yet, right?
Klaus:
I would say, if — I was in a meeting a while ago, a month or two ago, where a CIO stood up in front of the group and said, well, there aren't really any European relevant alternatives to the American hyperscalers out there. And I was like, why have we not talked to each other? You know, we think we've talked to a lot, but I think stop thinking that there aren't, because there is.
Jose:
So in your experience, if someone says that there aren't, it's not because they have done a thorough evaluation and seen that they're not, but more that they haven't, and they're ignorant about the possibilities?
Klaus:
I hope so. For us, the journey has been, in 25 years — a little bit, Europe was saying "Houston, we have a problem too, there aren't any relevant alternatives," to a phase where everybody is shouting from the rooftops "we have European cloud." And that's, of course, if you're a CIO or whatever, confusing, because you have to invest a bit of time in understanding, looking under the hood, who actually has a cloud and who has something else, and then finding out what do you want to use out of those different options. I see a lot of companies and public sector being in that space right now.
Jose:
You mentioned — and I'd like to ask about this — you mentioned the sovereign AI and the release in February. One thing you didn't mention, but it was 10,000 GPUs that you also brought in as a starting point. Is there any other milestone, or big, maybe call it a jump, that you've seen with these systems in the last year? Is there any other thing, could be an investment or some other capability that you've brought in, that you're proud of or that you see making a big change?
Klaus:
So I would say that was, I mean, that was a billion-euro investment for phase one. The EU's ambition is to have five of these AI gigafactories with roughly 100,000 chips. For us, we take it step by step. There's a question of market demand and catering for that, of course. So that's a massive investment. We're significantly investing more now in our European public cloud, because the demand is there at a different speed than it was before. So that's positive. So expanding where we have those. And then we're also making capital investments into different companies, into Gen AI, etc. And then, of course, we're embracing it violently ourselves. So one area we have invested is, you'd say, Gen AI Scrum Teams, if you want to call it that, to help with our own application modernization journey. And the beauty of being part of a telco is, telcos think in products. That means we don't just apply it to do a project and then put it away. We say, "How can we productify this so that we can actually also offer it in the market to our clients?"
Jose:
And you said AI Scrum Teams. So does that mean that you have a Scrum team only with AI agents running? Or what is the... Can you explain the concept of an AI Scrum team?
Klaus:
So, AI agents that have, like in a Scrum team, you have different roles, right? So anything from a product owner — of course, it's people and agents together, it's not either/or. But where you, as a human to your product owner, talk about or write about what is the desired outcome, what is the problem you're trying to solve. And then you have your agent, which can be different agents, depending on if you're shaping what you want to build, if you're the developer, if you're the tester, if you're the QA, different roles. And then you can let that run until you have something where, when you then look at it with a human eye, saying, okay, that looks quite okay.
Jose:
Interesting. Yeah. Thank you for sharing. And to wrap up, I would like to just ask a couple of quickfire questions. And one of them would be, if you have any thought leader or resource that you would suggest people hook into, or listen to, or read within this era.
Klaus:
So I would say my former colleagues — sorry for pointing to a competitive podcast — but my former colleagues Dave Chapman, Esmee van de Giessen, and Rob Kernahan have a podcast called Cloud Realities. They come around a lot of different topics that I think are interesting and valuable.
Jose:
Very cool. Is it based in Denmark?
Klaus:
It's a global podcast. Cloud Realities, yeah. And off this topic, but we all work in global or international organizations, and I would say a book to read, or somebody to follow, is Erin Meyer from INSEAD.
Jose:
Yeah, the Culture Map, exactly. It's so easy to forget. It's funny, I just suggested and lent the book to one of the colleagues just a month ago. So it's very nice to see that reference. Both her Culture Map, but also the one that she co-wrote on Netflix, the No Rules Rules. They're both super interesting as well. Thank you for sharing. Final question. To you, scalability is?
Klaus:
For me, scalability is to be chosen where it makes sense for you as a business. But I think we can fall in love with different concepts, right? It's about using it for the right purpose, I think.
Jose:
Perfect. Thank you so much for your time and for being here.
Klaus:
Thank you for having me. It's been fun. Thank you.
Jose:
And that's it for this episode of the Smooth Scaling Podcast. Thank you so much for listening. If you enjoyed, consider subscribing and perhaps share it with a friend or colleague. If you want to share any thoughts or comments with us, send them to smoothscaling@queue-it.com. This podcast is researched by Joseph Thwaites, produced by Perseu Mandillo, and brought to you by Queue-it, your virtual waiting room partner. I'm your host, José Quaresma. Until next time, keep it smooth, keep it scalable.
[This transcript is auto-generated]
Â