BARCELONA—Amid the flurry of sessions and briefings this week at Mobile World Congress 2025 in Barcelona, RCR Wireless News snagged a few minutes Karri Kuoppamaki, T-Mobile US’ SVP of technology development and strategy.
Here’s what we asked, and how he answered. This interview has been edited lightly for length and clarity.
RCR: How is the NTN beta going?
Kuoppamaki: Things are going well. We’re still in the middle of building the constellation, but we have launched over 500 satellites so far, which is pretty respectable, considering that we only started just over a year ago, launching satellites—that is due to our amazing partner, Starlink, and their ability to do those types of things.
… We have a large number of customers in beta now, trying it out. We’re getting a lot of good feedback. And I think one of the key learnings, to me at least, has been that it’s not necessarily intuitive for consumers that there is a difference between a satellite service and a terrestrial service.
They’re not meant to be the same. They serve a different purpose. We have large geographies in the U.S. covered by no one today—the mountains of the wilderness, the national parks and so on—but it is important to be able to provide the basic connectivity. But [people are] also used to using the smartphone the way we use it, and that creates an expectation that it works exactly the same way. And that’s not the intention necessarily, at least initially. That has been visible in some of the feedback.
But it works like a charm. Once you’re connected to the service itself, you can send and receive messages very quickly. We’re making good progress in looking into the next phase of capabilities, starting with maybe picture messaging, some light data after that. And then eventually our goal is to get to voice as well.
People are very appreciative of being able to stay connected while out and about. … But it also works the other way, and I’ll give you an example in my personal life. My son likes to go hiking in the backcountry. And I always worry, because there are avalanches in the Pacific Northwest. Just knowing that he can send a message to us makes me and my wife less worried, that we know he is okay. And that peace of mind is worth its weight in gold.
RCR: What do you see happening with data center evolution, and how does that relate to the network?
Kuoppamaki: I think cloudification is the foundation for everything, and part of our vision for the future is to move away from the way that networks were built in the past, into this fully cloudified, AI-native, multi-purpose horizontal networks that rely on distributed data center architecture.
We already have a very distributed network as it is, which is a great foundation for doing that. But we’re looking into and building our own cloud capabilities that not only provide services for ourselves and our internal needs, but also then become a platform for AI and associated services and use cases down the road.
There are multiple different ways of looking at AI: it will help us with our own internal AI workloads, help us with network insights and optimization, and provide services that result in improved customer experience. But eventually, you could also see us—the operator community—being sort of a service provider for AI-related services, at least on the infrastructure side, [for] different enterprises and potentially even consumer use cases down the road, to tap into some of those resources that reside in the data center.
RCR: How is T-Mobile using AI in a network context today?
Kuoppamaki: It’s all about customer experience. It’s two things: To get an easier way to get customer experience insights, as well as actions associated with that. … And then the other way ’round is that we can use AI to optimize things for us … but that is kind of the next phase of that evolution.
But there’s tremendous potential in harnessing the power of AI to make our lives as engineers a little easier: instead of looking at complex data and trying to find the answer in different documentation and manuals, AI has the power of translating them into something that is easier to understand and easier to translate into, what is the next best action in not just correcting things, but improving things for the benefit of our customers.
As a company, there’s three things that … we built and that I’m very proud of, which is the best network, the best value and the best experience. And especially in this best experience part, AI plays a very important role in taking that capability to the next level.
RCR: How is T-Mobile US thinking about Open RAN?
Kuoppamaki: We’ve been a believer in open interfaces for many, many years, even before Open RAN started. But the best way forward when Sprint and T-Mobile came together was to integrate the two networks using a more classical architecture, but we do use Open RAN in specific use cases as well. The way I see it is that the next evolutionary step in Open RAN is AI-RAN, which again, will be relying on some of those open interfaces but it is … more cloud-native and AI-native in nature.
Open RAN, as a concept, started before AI became a thing, so none of that is baked into Open RAN as it stands today. But AI-RAN has the potential of taking Open RAN to the next level. You probably saw our announcement [in September 2024] of the AI-RAN Innovation Center with Nokia, Ericsson and NVIDIA, and that is certainly part of that story for us down the road. So we are in the midst of that evaluation and development with our partners. It’s going to take a little bit of time. It’s not today—but if it all pans out as we expect, that will certainly be an option for our network evolution in the future.
RCR: How is T-Mobile US commercializing network slicing?
Kuoppamaki: It all starts with the best network in the nation, the largest 5G network that is Standalone based. Without Standalone, then it is going to be very hard to do well, [and] impossible to do network slicing. We started our Standalone journey in 2020, so we’ve been on that road for a long time—and that, combined with our spectrum assets and how it’s deployed in the network, gives us that great ability to start doing differentiated services for use cases that require that. It’s not just like one size fits all, you can actually do tailored things.
Sporting events are a great example of that, we’ve done a lot of proofs-of-concept and we know that there is demand for those types of services. So I think that is something that we will look at commercializing. We’ve also launched T-Priority for first responders, which is network slicing as well … and then also security, our SaSE-based solution is also something that benefits from slicing, because then you can remove the dependency of having a device-based configuration for that specific use case.
We have lots of different things under study and being evaluated right now, but again, the whole point is that if we didn’t have the network that we have today and the leadership that we have today, we couldn’t do what we are doing. And I think you will see over the next coming months and years, us really going big on services that rely on network slicing, because that demand is growing.
But again, technology is complex. You have to educate customers, you have to understand first of all what their needs are so you can map that to specific slice configurations, and then you also to have conversational dialogue with the customer base, so that they understand the benefits of slicing for their specific needs. And those types of discussions are also happening. Our T-Mobile for Business and T-Mobile for Government teams are doing a great job in having conversations and creating a pipeline of new types of services, and on the network side, we have all the building blocks in place. It’s just a question of saying, okay I understand your needs, so let me work on that.
