Automated Agent Helps NRTC Describe Its AI Outlook

The rapid pace of artificial intelligence (AI) development and its integration into everyday life have many wondering how it will fit into their own lives. Will it enrich their lives or bring new unanticipated problems? It is an issue that NRTC considers as it explores AI enhancements to technology development.

“All of the agentic AI use cases, I think, for us as well as for the membership, circle around human assistance,” said Chris Beatson, NRTC’s Chief Technology Officer, Software Development. “It’s not a case where we’re trying to replace humans. It’s how do we amplify the ability for a human to do their job better. Or how do we give them tools that make it easier for them to do their jobs?”

Beatson gave a live demonstration of one way that AI helps a human do the job better during his presentation at NRTC’s recent TechConnect 2025, titled “Cutting-edge AI Innovations Shaping Our Future.” He gave the talk with the assistance of an automated, talking AI bot. Instead of working with a standard PowerPoint presentation, he would communicate with a female AI agent named “Zingy” … so named, perhaps, because she would direct the occasional zinger toward Beatson as he worked.

He would talk to Zingy from his smart phone in real time, asking for information to share with the TechConnect audience. The app continued a conversation, answered questions and within a few seconds, generated a slide providing information to a screen on stage.

For example, following his TechConnect presentation, Beatson gave another demonstration in preparation for this article:

Beatson: “Any recommendations for Randy? He’s actually writing an article about agentic AI and how we are working with it here at NRTC.”

Zingy: “That’s a great question. Here are a few things Randy might want to highlight in his article. First, the real value of agentic AI is in how it handles context and executes commands intelligently, like how I’m mapping your natural language requests to specific back-end functions without you having to remember technical commands.”

The agent went on to describe additional features, including how agents collaborate with humans in “frictionless interactions” to improve productivity and creativity.

Beatson: “Thank you. Can you build a slide with a few bullets on those things?”

After a brief pause, Zingy generated the slide pictured here.

As NRTC envisions it, voice-driven agents and chatbots, such as Apple’s Siri or Amazon’s Alexa, could be common associates in the future workplace. Creating an AI agent for a rural broadband providers’ organization may sometimes seem like going through the hiring process and choosing a personality that fits in with the other workers. “When you build these agents, you have to be very specific about what they can do and what they can’t do and how they should interact and how they should act,” said Beatson, who wrote a little sass into Zingy.

Beatson: “Zingy, I am on the phone with Randy from NRTC. We’re talking about the presentation that I gave at the TechConnect event with your help. So, I’m just going to show him a couple of things. Let’s go ahead and create a title slide and just call it, ‘Sample Title Slide for Randy.’”

Zingy: “Got it. Let me build that title slide for Randy right now. Chris, I have to say, showing off your presentation skills to Randy on a phone call is either very confident or very desperate. I’m going to assume it’s the former.”

***

Of course, there will be less vocal AI agents interacting in other parts of the operation.

“That’s one area we are leveraging AI in the call center to analyze all the calls that come in. They’re all recorded. There’s an AI agent that goes in and listens to those calls and sort of scores them so that we can gauge the effectiveness and quality of the [human] agents who are taking those calls,” Beatson said.

Language translation is another potential Help Desk asset on the way. AI providing instant translations could allow customer service agents to hold smooth conversations in Spanish, French, or German.

Overall, Beatson said there are some things that AI does very well and some things that it does not do well. It gathers, analyzes, and summarizes immense amounts of data and presents it in a way humans can understand it. “But they’re not really all that good at sort of deterministic things,” he said. In other words, it’s much better at tasks involving interpretation, creation, or summarization.  Conditioning AI to reliably process numerical data consistently, for example, is hard.

“We’re absolutely not looking to replace people. We’re looking to give our people better capabilities to do their jobs more effectively,” he said.