How Extreme is Scaling AI for High-Density Venue Networks

Extreme Networks’ CTO EMEA Markus Nispel spoke at Mobile World Congress 2026 in Barcelona, setting out how a new approach to networking aligns with telco operators’ need to manage high-demand environments.

Platform One, the company’s AI networking platform, revolutionises connectivity with agentic AI. The systems act with a level of autonomy while remaining under human oversight, and are built to scale across large estates of devices and users.

“Platform One was designed and created from scratch to support AI and agentic workflows specifically,” says Markus Nispel, CTO EMEA at Extreme Networks. “This really starts from bottom up with a clean data architecture.

“We re-architected our entire data pipeline end to end to feed into AI effectively, then scaled AI on top of it like a software engineering exercise.

This allows for the management of “a couple of million network devices and tens of millions of endpoints”, harnessing years of cloud experience for seamless AI scaling.

Stadium connectivity tests telco capabilities

Stadiums are high-density venues and present a perfect clear test case for telco infrastructure.

Wi-Fi and 5G networks often struggle when tens of thousands of users attempt to connect at once, which causes interference and bandwidth bottlenecks.

Extreme Networks began working with stadiums in the US in 2013, mainly with the NFL, before expanding into Europe. It now deploys its connectivity solutions at Old Trafford, the home of Manchester United.

“It’s a combination of art, science and experience to serve 50,000 or 100,000 fans in a stadium – it’s not easy,” Markus says.

“The design and architecture is critically important. We’re using AI to optimise the RF environment and to deliver a unique experience to fans in the stadium itself.”

AI-driven optimisation helps adjust RF (radio frequency) signals dynamically to reduce interference and improve coverage. This reflects similar techniques used in mobile network optimisation, where software continuously tunes performance.

The result is consistent connectivity across entire venues. Fans access high-performance Wi-Fi from car parks to turnstiles, supporting digital ticketing and reducing delays at entry points.

Fans wanting to share on social media and stream video rely on stable connections, which AI provides.

“Teams want to connect to the fans. They always say that when you’re in the heart of what that team is all about and you can’t communicate with your fans, that’s a problem," says Markus.

“Fans obviously also want to share their experience outwards, so that’s a multiplier as well.

“This means that all these venues and teams are rethinking their digital strategy – including the stadium experience.”

This shift aligns with broader demand for enhanced user experiences from telco providers. Connectivity underpins services, engagement and revenue opportunities, with data at the heart of telco strategy.

“It’s all about the data that the network is producing,” he says. “Optimising network performance is one thing, but creating and generating insights and analytics around fan behaviour and optimising offers towards those fans is another.

“Data allows us to optimise the flow of the crowd within the venue, for example, and also you can create new services and offerings on top of that.”

Security and segmentation in dense networks

As networks scale, security risks can also heighten. Stadiums contain thousands of users alongside operational systems such as media infrastructure and venue controls, and this creates a complex environment with many potential vulnerabilities.

“Stadiums are tricky because they are dense environments,” Markus says.

“The technology we’ve deployed in the last few years allows us to segment the network infrastructure effectively – because you not only have fans at the stadium that you’re connecting, but you also have the operational technology of the stadium itself from press to other various teams.

“Segmentation has to be able to integrate with security monitoring tools for threat response. That is our solution to that problem.”

Alongside security, trust in AI systems remains a concern for operators and enterprise customers alike.

“Explainability is critically important as we look at using AI for network operations,” he says. “Trust drives adoption, and you can only gain the trust of your users using those tools with transparency.

“Within Platform One, we expose all of the reasoning and planning that our agents do to our users so they can really understand how certain decisions are being made.

“This allows the user to action their own approval in a human-in-the-loop type of setting. This means humans are always in control, but with the knowledge and the transparency on how an agentic system came to that result.”

Telco lessons for AI deployment

The adoption of agentic AI begins with practical use cases rather than broad strategy. Markus frames this as identifying where AI can create new services or improve existing ones.

He says: “You should think about how to use AI to create new experiences. It’s not just about automation and efficiency, but how can you leverage the technology to do something unique.

“Expose your employees to it, let them play around with it. Once then you have identified what you want to go after, ensure that you assemble a cross-functional team and try to deliver a capability end to end.”

This reflects a shift from experimentation to deployment, and Markus encourages telco operators already managing complex infrastructure to integrate these systems into live environments.

“Don’t just talk about AI, experience it,” he concludes.

“Use it in your private life and your business life alike because we’re at a huge inflexion point that is changing not only the networking industry, but society and humanity.”