Turning 5G Into Revenue With AI, Automation and RAN Data

Hi everybody. My name's Linda Hardesty and I'm Chief Analyst for Communication Technologies at Fierce Network. I'm here today with Shaun McCarthy, he's president and chief revenue officer at Spectrum Effect. We're going to talk today about the challenges that operators face in maximizing 5G investments, the growing role of automation and intent-based networks, and how emerging approaches like Spectrum coexistence and AI-driven optimization are shaping the future of mobile networks. Welcome, Shaun.

Hey, Linda. Thanks for having me.

To kick it off, for listeners who may be new to Spectrum Effect, how would you describe what you do and your role in today's networks?

Spectrum Effect was founded by industry experts in the RAN space, radio access network space. And they went out to tackle what they viewed as the most challenging problem in the RAN, which is really interference and interference's impact on the use and utilization of Spectrum. So the company had its learnings over the years, and really I'd say in the last two years or so, really hit its stride with deployments in all three US tier one operators that we completed last year. And basically what we ended up building and the product sort of evolved to is a platform that uses the RAN as a sensor. We use machine learning and AI algorithms to do analysis and look at the data that we get from these sensors coming from the RAN. And then we drive some sort of actions back into the network, whether it's providing insights or doing closed loop automation while we're reprogramming the RAN to optimize a user's experience.

So we kind of started there, really focused on RAN optimization. And over time we've sort of evolved as customers have said, "Hey, you're looking at this, you've got this sensor network out there, you've got these AI models. What else can you do?" And we've kind of evolved some new applications around Spectrum Sharing and we're in the process of evolving these new applications around some of these enterprise 5G use cases that are going to leverage the macro network instead of private 5G networks.

Wow. Interesting. The RAN as a sensor. Okay. And so AI is everywhere right now, as we all know. How do you see telcos approaching AI today?

Yeah. So AI is certainly... You can't turn without hearing that acronym. I think about AI for telco really from two lenses. One is AI for the telcos themselves. So that's pretty obvious. Like every company on the planet, all of the telcos are looking to integrate AI to every aspect of their business, from HR to operations, to internal operations, sales and marketing, the chatbots when we call into these call centers. But obviously they're rethinking how they use AI to build and how they use AI to operate networks. And so that's kind of the space that we play in. And there's really two areas that we're pretty active with operators. One is there are these really important engineering teams that are keeping these networks alive, that are optimizing these networks, that are evolving these networks to customer requirements. And AI can deliver insights to these folks to let them do what they do in a much more effective, efficient, and highly scalable way. So one is using AI to deliver insights and enable your people.

And the other is there's really a lot of things that people are doing today that could be automated. And so with our solution, we're optimizing a network. And so rather than tell the engineering team, "Hey, this is how you should optimize the network, or this is the problem you have in the network," we can just go reprogram the network ourselves. And so that's really where you get the closed loop automation to drive resolution of these issues and to free up... Your people are your most important resources. So free them up to go work on other higher order problems. But the other is about telco for AI. And so that's really about how are operators going to take advantage of this new wave and how are they going to monetize this wave of AI and the new use cases that are coming from this AI movement.

And I think telcos have a really great opportunity and I think we're still really early in the process. So the one that everyone's talking about today is this AI RAN concept and we're distributing AI inferencing at the edge using the same hardware for the actual RAN as we are going to use those GPUs when they're not active in the RAN and how we're going to build AI into the RAN itself. But the other one, like I mentioned, is we're seeing these 5G use cases evolve around AI that are sort of ... I think it's what we talked about with 5G years and years ago, but it's finally coming to fruition. So we've got this huge sensor network and as enterprises are looking to turn up fleets of drones and vehicles and robots that are going to interact with the mobile network, you start to ask yourself, what are the things that operators can do to help monetize these services and how can we help them do that and also deliver value to those enterprises? So it's a really exciting space right now.

Yeah, for sure. Well, many operators are still struggling to fully monetize their 5G investments. From your perspective, what are the key challenges in realizing that value at this stage?

Yeah, it's funny, right? So before we launched 5G, as an industry, we said, "Hey, 5G is going to be all about the enterprise. It's all about the enterprise." And then we poured billions of dollars into these macro networks, which are largely built to deliver consumer services today. And we didn't really deliver much more for those consumers, therefore the consumers didn't want to pay more. And we all looked at ourselves and said, "Hey, this is not a good business case for 5G." But listen, I think part of the problem is all of these initial enterprise 5G use cases, which are really compelling, they're all being served today via private 5G networks. So warehouses and mines and factories are all deploying 5G and these use cases are real and there's strong value there, but the revenue for those use cases isn't correlated with the investment we made in the macro networks.

But like I mentioned, I mean, we're still early, right? We're still early and we're seeing more and more of these use cases start to get turned up where these enterprises are going to be leveraging the macro network. So like I mentioned, think about fleets, fleets of drones, fleets of vehicles, fleets of robots, right? These are devices that are going to live outside of this boxed-in private network. And so we're seeing that happen. So I think it's not typical of us in the telco space to get really, really excited about a technology and kind of give ourselves some unrealistic timelines. So I think it's all coming to fruition and I think operators are going to continue to have opportunity to grow their revenues based upon the 5G use cases that were promised. It's just coming a little bit later than I think we originally envisioned.

A lot of AI in networking is still discussed in terms of trials and pilots. What's been most challenging about moving AI into the actual day-to-day network operations?

If you go back to really 2024, which was probably kind of the breakout year for AI, it was really a year of excitement, but also a year of analysis paralysis. We were really over-engineering these business cases and so much energy was spent on these trials and what's the ROI of this and what's that? And part of the problem was, at least from my lens and from my experience in the space, there was a lot we didn't know. So if you don't know what you don't know, how do you build this business case? And so what we learned is when we started doing deployments, we were able to solve problems with our deployments and with AI that we never would've imagined it if we purely just did that ROI beforehand. And we did, of course. And so the ROI got much better after the deployment. And I think the tone changed quite a bit in 2025.

So we transitioned a bit from over rotating on the business case to folks saying, "Hey, look, we just got to get moving. This is clearly the direction we need to go. How do we move? How do we go faster?" And I think that's kind of where we're at now. It's really about scale and speed of deployment, and it's not about trials. It's more about, let's get something deployed, let's iterate on it and let's scale it. Really that's kind of the mindset I think going into 26. So I think we're past that sort of over analysis paralysis phase.

And that's been a little bit of a culture change for Telco, huh? Let's just do.

It's a huge culture change. It's a huge culture change.

The industry has talked about automation and intent-based networks for a while now. Where are operators actually making progress today and where are they still struggling?

Yeah, it's interesting. I think about this one, and I think it's quite analogous to self-driving vehicles. The answer is obvious. It's inevitable we're going to move into this direction, but we're always thinking about what could go wrong. And it's kind of counterintuitive because I think we all know deep inside that machines are more accurate and they're less error prone than humans. But you don't want to be the person that gets into an accident because you were letting your car drive itself. It's kind of an embarrassing thing. And I think networks and the operation of networks are very similar. No one wants to be on the front page of the Wall Street Journal because there was an outage and there's an FCC violation because of a mistake from this automated network operations. But like self-driving cars, it gets better over time. So we get more and more comfortable letting machines take the wheel.

We're more comfortable. It's not unusual to see a car, self-driving taxi if you're in San Francisco or someone's Tesla gets summoned to go pick them up outside of a store. So it becomes more of the norm and we get more comfortable with it. And that's kind of where we're at with network operators today as well. They're now taking their hands off of the steering wheel. Over the past year, they took their hands off the steering wheel and I think the tide's turned quite a bit. So before we were afraid of messing up because we were using closed loop automation and tent-based networks. Now we're afraid of getting left behind because we haven't moved in that direction. We're trying to do it the old way and we can't move fast enough. We don't have the right economics to compete with our competitors who are automating things.

And so that's kind of where we're at now. I think we've crossed this tipping point and it's really about what other use cases can we automate? Where else can we do closed loop automation? Where else can the networks drive themselves and be autonomous versus needing humans to control everything?

All right. Let's talk a little bit about Spectrum Coexistence and ISAC. How do you see these concepts evolving and what role should industry and/or government play?

Definitely a hot topic in the US specifically. But I mean, ISIC in general, which is integrated sensing and communications, it's pretty broad. It's going to be a big part of the 6G conversation, but it's not limited to 6G. It's really thinking about sensing, integrating sensing and communications. And so naturally when SpectrumNet is sort of the global leader in this RAN as a sensor, we've created this space of using the RAN as a sensor platform. We're right in the middle of that conversation. So there's going to be a lot of interesting uses coming out of ISIC and all the working groups around ISIC.

One of one where right in the middle of, we actually had a breakout year in 2025 is around Spectrum Sharing or Spectrum Coexistence. And generally when we talk about that, it's sharing between the US federal government and the mobile operators. And the problem we're trying to solve today is the mobile operators want... Need, not really want, they need more spectrum and there isn't really a healthy pipeline of Spectrum available to them. And some of the best Spectrum is in use by the federal government. And so how do we allow these two things to coexist? The federal government can't be too costly time consuming for them to move off of it. How do we allow these things to coexist? And if you think about the way that would operate, the fundamental processes, the federal government, they're the incumbent, it's their Spectrum to use. If you can sense when they are or not using that Spectrum, and you can then trigger, you can then take that data, that insight back to the operator and say, "Hey, in these parts of the network, the incumbent is active. You can't use that Spectrum. In this parts of the Spectrum, the incumbent is not active, you can use it."

So you can imagine that this gets very technical, gets very fast, very dynamic, and it's actually exactly what we initially built for interference. We sensed that there was interference in the network, we thought and figured out what the best thing to do, and then we reprogrammed the network to give the best user experience, the most throughput for the operator. This is the same exact use case. It's just instead of looking for interference, we're looking for incumbent radar signals from the federal government, and then we therefore trigger the operator, say, "Hey, you just got to kind of maneuver around that." So it allows you to very granularly maneuver around the spectrum. So it's not like, "Hey, shut it all off." It's more like, "Hey, they're here. Just don't use that little piece of it."

And so we had a really breakout year and we're really excited about that. We're working with a mobile operator, tier one US mobile operator on a project there for some Spectrum that they have. And then we also were part of a winning bid from the Nokia Federal team that won a project with the Department of War to do a proof of concept of this through the National Spectrum Consortium. So really hot space, really exciting. It feels good because you're doing something that's good for the nation as well. So it's a good space to be.

5G has always been targeted toward enterprise use cases. How does your technology play into that?

Yeah. So like I mentioned, I mean, all of the initial 5G use cases have been delivered via private networks. And maybe I'm exaggerating when I say all, but pretty much all of them have been delivered through private networks. And so that's, how do I automate my warehouse and have robots that are using 5G? And those are mines where we don't have self-service private networks. So that's how I'd say the first half of the 5G journey has played out, but things are evolving and AI is actually helping drive a lot of these new use cases. And so we're seeing a lot of activity, a lot of investment in these enterprise fleets, be it drones, vehicles, robots. And again, all of this stuff, there'll be personal consumer sort of versions of this as well, but think of big enterprise fleets. And if you think about those, these devices or... These devices, I guess, they're going to be exercising the mobile network in ways that it's never been exercised before.

They're going to be doing different things than what we do on our handsets. And so we're in a prime position to help operators build new revenue streams and enable new capabilities for these fleets. What we always get is, hey, you're now using, you've got millions of sensors that you're looking at constantly. And what if you put a different AI model in there and that AI model could look at it from a different perspective, what things could you do? And that's kind of where we're at today. So it's probably early stages of our kind of role in it, but collaborating with operators, then we're also collaborating with some enterprises in this space. And we think it's a really interesting three-legged stool to deliver a lot of value for the operators as well as for the enterprises.

Okay. What trends and priorities do you see emerging in the telco industry this year?

Telco has been challenged, I guess after the COVID boom, boom kind of ended, we went through a bit of a rough patch, I'd say, in telco. And I'd argue that 2025 was a bit of a breakout year for telco. And I only breakout, I mean, a lot of the telco ROI, the business metrics aren't where we want, the growth isn't where we want to see it, but in terms of the technology, it's kind of gotten pretty exciting. There's this whole interesting terrestrial, non-terrestrial thing that coming together as we're looking at the satellite space. The AI RAN thing is obviously pretty cool and pretty exciting. And NVIDIA is part of the RAN conversation. It gets kind of exciting. We're seeing more and more use cases and exciting use cases that are going to leverage 5G in more interesting ways. So I think you're going to continue to see that evolve in 2026, and you're going to see some changes to the way we build and operate networks.

So one example you might think about is today networks are built with an 80% downlink, 20% uplink. That's based upon our mobile phones that we use. That's what exercises and taxes the mobile network today. Well, when my glasses are constantly uploading data, when my pin, my AI pin is constantly, you have a very different profile of uplink to downlink. So there's going to be some pretty interesting changes that's going to really have us rethink a lot about what we do. And AI is going to continue to be the theme. But again, I think I feel like we finally like the cool kids again being back in telco, maybe not as cool as the AI kids, but telco is kind of getting exciting and there's some really interesting technology problems that we're tackling.

Yeah. Thank goodness for AI and satellite, you're right. We're back in the saddle.

Yeah. Right.

Well, we're starting to hear about 6G more frequently. Given the state of 5G, how should we be thinking about 6G?

Yeah. 6G, I think, listen, I think it's a bit of a loaded question. I mean, we just talked about how the 5G ROI has been challenged and I question what type of appetite are the operators going to have to go spend billions and billions of dollars to go refresh a bunch of hardware? That's going to be a tough one for operators to take on. At the same time, the Gs are really important. I mean, the industry rallies around these Gs. We have this one common north star that we all rally around. And I think that's pretty important. It's pretty critical for our industry. So yeah, I mean, I think, listen, 6G's going to happen and there's going to be some cool things we're going to do in it. And ISEC, the integrated sense and communication is going to be a big part of it. I just really think it needs to be a big focus on software.

What are we doing with software? A lot of the AI that we do today is really kind of above the network, if you will. I think you're going to see the AI like the stuff that we do as well as other AI type capabilities get kind of pushed more into the network and become more part of the network. But I think the key thing for the industry is we got to really be thoughtful about the business case and the monetization upfront. I mean, I think for operators to continue to invest for them to invest in 6G, it can't be, "Oh, there's going to be enterprise use cases out there." We need to really have a good understanding of where we're investing the dollars, we're going to get a return on investment in those dollars. And I think we're going to figure that out. Like I say, I think we're going to get a lot of momentum in the back half of 5G as well, which will make a lot of us feel a lot better about going into 6G.

But generally speaking, I think it's a lesson learned that we need to probably be a bit more thoughtful upfront. But again, we'll see what happens. It's still kind of early in that conversation. Now we're starting to get the early buzz, but I'm excited to see how it all plays out.

Okay. Well, I think that's all the questions I have, Shaun, but I really appreciate you chatting with me today and thank you for being on Fierce Network's podcast.

Great. Thanks for having me.