Start-up NetSpeek Uses AI To Make Managing Big Screen Networks Easier

January 21, 2025 by Dave Haynes

A start-up with roots in PC-based digital signage media players has launched a unique new service aimed mainly at pro AV solutions providers and end-users – using artificial intelligence to largely automate and streamline the management and integrity of enterprises that use a lot of screens and other devices.

NetSpeek, which comes out of stealth mode today and makes its public debut in a couple of weeks at ISE, represents two years of work for Erik DeGiorgi, who many industry people will know as the guy who was running the specialty PC and services firm MediaVue Systems up to its sale early last year to UK-based Assured Systems.

The core idea of NetSpeek is that Large Language Models can speed the turnaround time on supporting and troubleshooting AV solutions running across enterprise-level networks, and automate many of the tasks needed to ensure equipment is ready for everything from video-based meetings in conference spaces to signage around an events facility that is steadily changing.

NetSpeek’s initial sales and marketing focus is on unified communications and the other AV technologies commonly found in corporate and educational environments, which makes both because of scale and complexity. The digital signage ecosystem has several companies with good in-house or third-party device management solutions designed to maximize uptimes, but DeGiorgi still sees a roll for NetSpeek complementing what’s available and on place.

DeGiorgi says in his MediaVue days his team consistently heard that while individual system and device management tools worked well, organizations struggled because numerous tools were needed and information was scattered.

“Nine out of ten customers told us they had too many management tools for discrete vendors,” explains DeGiorgi. “We realized we don’t want to build another management tool – we wanted to build a platform driven by generative AI that can handle day-to-day operations autonomously.”

That insight led to a two-year development process by a small team in Boston, resulting in a platform that can now manage mixed-vendor AV environments.

At the core of NetSpeek’s platform is LENA (Language Enabled Network Administrator), an AI system built using what’s called an agentic architecture. Unlike traditional monitoring tools that simply provide device status, LENA involves multiple specialized AI entities that work together to manage different aspects of an organization’s AV infrastructure. These entities include specialists for various systems like Microsoft Teams, Zoom, and different hardware manufacturers.

The platform can handle common tasks that typically require manual intervention.

For example, it can automatically run room checks before meetings, testing conferencing equipment, displays, and other AV components. If issues are detected, LENA can often resolve them automatically – such as logging back into disconnected services – or alert maintenance staff about problems like unplugged cables that need a physical visit. Fixes can get prioritized based on scheduled meeting times.

For system integrators and IT departments, NetSpeek is pitching that it offers efficiency gains in installation and maintenance. The platform can generate step-by-step workflows for hardware installation and testing, potentially reducing the need for highly specialized (and expensive technicians) when  a still well-paid but more general AV tech can do the work for half the cost or less.

The premise of turning management of devices across a network inevitably conjures up worried questions about AI taking over, like Skynet in the Terminator movies. DeGiorgi says security and control were primary considerations in the platform’s design.

“We have guardrails in our system – we’re only leveraging internal information to conduct our work, not going out to the internet. LENA asks for confirmation before pushing any actions, unless you specifically authorize autonomous control.”

The platform operates within an organization’s internal network, with a containerized approach that helps maintain network security while still allowing AI to effectively manage AV resources.

Implementation can take different forms, depending on customer needs.

For new deployments, the company handles integration with various vendor systems and devices. The platform can also be integrated with existing management tools, adding AI capabilities to already-deployed systems. Most of the platform operates in the cloud, with local edge clients deployed on organization networks to interface with devices.

The business case for the platform is built on both operational efficiency and workplace productivity. “Based on our pricing model and costs, the platform pays for itself if you save one average salary corporate employee two minutes a day,” says DeGiorgi. “Every single person has walked into a conference room, sat down for a meeting, and had it not work. We can dramatically reduce that experience.”

On the operational side, it can help organizations handle increasing AV management workloads without proportionally increasing staff.

Inevitably, using AI and automating tasks and aspects of monitoring and troubleshooting raises the prospect of software bots taking human jobs in network operations centers. That could happen, of course, but DeGiorgi says everything he has heard from end-users and big AV/IT companies that manage large, complicated networks is that there now tends to be more work to do than people to do them. Rather than AI taking out 15 of 30 jobs in a NOC, it is more likely that AI will be able to handle most routine tasks and raise solutions to problems much more quickly – because it can do things like comb through entire trouble ticket archives in seconds, instead of minutes or hours. So those 30 NOC staffers have more, let’s say, quality time on their jobs.

While it might seem like the IT teams at big companies would be well down the path of using AI, DeGiorgi says his many conversations suggest it is still a roadmap item for most. Few have established formal processes for evaluating and deploying AI systems.

He’s been talking to AV integrators about attaching NetSpeek and LENA to their existing workflows and technologies. Part of the development work that’s been done to get NetSpeek to launch-ready has been integration with technologies common across AV systems, from unified communications to digital signage. NetSpeek includes, for example, the ability to monitor and manage various devices from manufacturers like BrightSign, Sony, and Samsung.

The company already has a platform-level integration with XYTE, a management system developed by former Crestron team members.