Erik DeGiorgi, Netspeek

The 16:9 PODCAST IS SPONSORED BY SCREENFEED – DIGITAL SIGNAGE CONTENT

The people who build and maintain very large networks of displays, PCs, servers and other devices tend to have more to do than time to do it, and when some technical shit hits the operating fan, trying to work out what's happening and what to do about it takes experience, brainpower and what can be punishing downtime.

So what if generative AI could be used by a network operations center team to comb through knowledge bases and trouble ticket archives to identify solutions in seconds, instead of minutes or hours? And what if a lot of meat and potato workflows done to deliver services and maintain uptimes could be automated, and handled by an AI bot?

That's the premise of Netspeek, a start-up that formally came out of stealth mode this week - with an AI-driven SaaS solution aimed at integrators, solutions providers and enterprise-level companies that use a lot of AV gear. The Boston-based company is focused more at launch on unified communications, because of the scale and need out there. But Netspeek's toolset is also applicable to digital signage, and can bolt on to existing device management solutions.

The guy driving this will be familiar in digital signage hardware circles. Erik DeGiorgi was running the specialty PC firm MediaVue, but sold that company about a year ago. Since then, he's been forehead-deep working with a small dev team on Netspeek. We caught up last week and he gave me the rundown.

Subscribe from wherever you pick up new podcasts.

TRANSCRIPT

Erik, nice to chat once again. You sold your company about a year ago, and I don't want to say disappeared, but kind of went off the grid in terms of digital signage, and now you are launching a new company called Netspeek. What is that? 

Erik DeGiorgi: Thanks for having me back, Dave. It's crazy. Time flies. I think it's well over two years at this point since our last conversation. 

We launched Netspeek at the beginning of the year. At the same time, we sold out MediaVue. Netspeek is bringing to market the first generative AI platform focused on supporting the day-to-day operations of mixed vendor estates of pro AV networks. Digital signage is certainly a component of that. We're really focused on the totality of pro AV technologies. So it includes a lot of UCC unified collaborations and communications technologies as well as signage, and really targeting office spaces. So think about meeting rooms and conference rooms. You might have a Zoom or a Teams environment in there as well as a signage system or classroom environments, and what we've developed is a generative AI solution that can be embedded into those networks, that can work alongside human operators, network administrators, technicians to help them support them in their daily workflows, and then also bring a large amount of automation.

So our platform can not only kind of observe what's going on in a network, kind of a 24/7-365 way, but then take action and use its own logic and reason and independent thinking to analyze situations the same way a human operator would and then structure and generate responses. So being able to directly address equipment and solve problems independently. We're pretty excited to bring that to market. We're launching to the industry here in a week, and then we'll be demoing at ISE at the beginning of February. 

You’ll have your own stand at ISE?

Erik DeGiorgi: Yes, and I did pull up the booth number ahead of the call, but of course now it's on a different tab. It's in the Innovation Park, and the booth number is CS820, and it's actually centrally located there in the Innovation Park. So actually right outside the digital signage area. 

Yeah, I think for people going to ISE, the Innovation Park is kind of along the main corridor in between halls. 

Erik DeGiorgi: Yep, it's the central hallway. 

Okay, so people should be able to find you there. 

Erik DeGiorgi: Hopefully, yep. 

Not a sprawling booth like a Samsung or LG or something, but… 

Erik DeGiorgi: We measure in single meters. I think it's a 2x3 meter booth. 

Startup life.

Erik DeGiorgi: The price was right. 

There are lots of device management platforms out there, either independent third-party platforms that you would subscribe to and bolt onto your system or a fair number of companies, whether they're integrators or CMS software companies in the context of digital signage have their own device management code written in, how is this different? 

Erik DeGiorgi: Yeah, absolutely. Netspeek is not another monitoring platform. Monitoring is a necessary component, right? You need to know what you have on the network and know what it's doing as a foundation. But our value really lies in the intelligence that we're bringing into that. So it's taking that monitoring and observation, but then actually doing something with it in doing that either again to assist a human by bringing kind of an encyclopedic knowledge and institutional knowledge or whether it's through the automation, and so we're going to market with a total solution.

We have a monitoring platform that we've developed as a necessary part of our total solution, but we actually are also partnering with existing remote monitoring and management platforms to essentially bolt on to them, and then bring that intelligence to their monitoring platform and actually at ISE, you'll be able to see that as well. 

So they should happily run in parallel using APIs or…? 

Erik DeGiorgi: Yep. So we hook into the existing monitoring platform and we essentially bolt on the, the reasoning and the intelligence, and then allow an existing user to leverage that front end, and that monitoring platform that they're already familiar with.

Who do you think you're primarily going to be selling this into? Is it like integrators and service providers who have network operation centers or would it be end users? 

Erik DeGiorgi: So it's a little bit of both, and candidly at an early stage, you tend to take a bit more of a scattershot approach, and test where the value emerges. It's a new technology, gen AI, everybody knows it's there and in a large part don't know what to do with it. But we've kind of honed in on three initial go to market opportunities.

One is like a total solution directed towards the end user. One is more of a channel centric focus, whether it's a system integrator managed service provider. We're actually already engaged with a few, of each, that are interested in leveraging the platform in that capacity. And then also, like I said, with, an existing management. You could be a manufacturer. So think about even an independent manufacturer, or a platform provider, like an existing monitoring platform. So an existing tool is specific to a manufacturer or a tool for more broad-based management. Like I said, we can kind of bolt into those and go to market that way as well. 

So in the scenario of a network operation center in the context of digital signage, an integrator that's doing the work to monitor a large QSR network for a restaurant chain that doesn't want to do that internally and they've got a whole bunch of screens up on a wall and they've got big curved desktop screens and the whole bit and they're watching what's going on. 

Is the idea here in part that. As a problem develops and it's kind of weird and not familiar that if you had to go into a whole bunch of manuals and archived information, it would take many minutes, maybe even hours to do it versus if this is all on a learned model that the solution or at least ideas on a resolution could come up in seconds?

Erik DeGiorgi: We really kind of lean into the personification of our platform. So our product is called Lena and Lena is an acronym that stands for Language Enabled Network Administrator. So we really have modeled the platform and the solution after the workflows that human operators perform every day. So imagine being in that knock and sitting there next to your colleague, Lena and Lena happens to be trained on every respective certification related to the deployment, and has been trained in every application software that's being used, has an encyclopedic knowledge of every technical document for every piece of equipment or technology that's in that deployment, has the ability to - at the speed of light - comb through any historical information like previous support tickets or anything like that that's been related. 

So being confronted with a situation, whether that's a critical situation or whether it's looking at something that's preventative, or maintenance-oriented, just imagine having this kind of superhuman user that can just as a human operator analyze the situation, develop a logic flow, think critically about that situation, pull in outside information to help diagnose a potential issue, construct a resolution, and then either autonomously or along with a human companion and approval, go ahead and execute that action.

One of the things that Lena can do out of the box is we've done all the integrations, and I say all we've done many, and we're continuing to do many more integrations with all the different devices and technologies that you see in these networks. So, as a generative AI, Lena can generate information for human consumption, but Lena can also generate structured information that translate down to device commands in various ways. So Lena can actually take action and do things on her own, and, we default to saying “her” because get used to personifying. Some people lean into that, s