Scout InsurTech Interview with Wess Anderson

Wess Anderson is the Co-Founder and President of Agency VA, a company that helps insurance agencies put agentic AI to work alongside their existing systems and staff. Wess was interviewed by Chris Luiz, Co-Founder and CEO at Scout InsurTech, to discuss how agencies of every size are starting to adopt AI, what actually has to change in their tech stack to make it happen, and the concrete first steps any agency owner can take today.

Wess, you've talked to hundreds of agencies about adopting agentic AI, and the same question keeps coming up: "I know I need to start, but where do I start?" What's your answer to that?

“What we're seeing in the AI world is that these various AI platforms coming out have a lot of similarity. It's not as if any one agency is going to get some bleeding-edge new AI that nobody else has access to. The place where you actually have to start, and the greatest separator when it comes to AI, is the data that the artificial intelligence is learning from. So step one for any insurance agency is making the data you already have access to AI enabled. That's really the first thing we do with agencies: we get their data set up so artificial intelligence can learn from it.”

You've mentioned an interesting contrast: an 18-year-old in high school managing eight agentic AIs and nearly $1M in premium, alongside multi-billion dollar agencies doing something surprisingly similar. What do these two examples tell us about who should be adopting AI right now?

“That's a great case study, and it's a live one. We're watching it happen in real time. The 18-year-old is a senior in high school, still taking classes, very tech forward and tech driven. They've been commissioned to manage close to $955,000 in premium, and their job is managing eight agentic AIs. The retention numbers and growth numbers we're seeing from this individual are beating the benchmark norms of traditional agencies, even though they're a rookie who only recently got insurance licensed. On the other end of the spectrum, we have clients whose hundreds of agencies together are worth billions upon billions of dollars, and they're doing the exact same thing: using a human to manage multiple agentic AIs to increase the value of the agency, improve retention, and drive growth. That spectrum is so wide because the group of people who need to be adopting AI right now is that wide. It's everybody. It's not isolated to just the very large agencies, and it shouldn't be limited to the young, early adopters either. Every agency across that spectrum should be using AI inside their operations.”

A lot of agencies hesitate to adopt new technology because they're afraid of what they'll have to tear out or overhaul first. What actually needs to change in an agency's existing tech stack to start using agentic AI?

“I think that fear is valid, especially given the technology of the past. In the olden days of insurtech, adding a new piece of technology could mean pulling out the alternator, pulling out the engine, and putting in a new one. That's a terrible process, and nobody wins. Agentic AI works differently. The agencies we work with have data sets living in AMS platforms, in IVANS, moving through carrier sites, and sometimes in a separate CRM on top of all that. The thing about artificial intelligence is that you don't pull any of that out. The agentic AI sits amongst all of it, and you get the efficiencies of AI without disrupting or getting rid of your AMS. Once you get that AI enablement in place and the data structured properly, the agentic AI works on top of and with all the tools that already exist. Nothing has to be pulled out, and no existing workflow has to be disrupted.”

Once an agency sets up its first agentic AI, what role does a human still need to play? And realistically, how many agencies already have someone who can fill that role?

“One of the things we've done well at Agency VA is train and manage humans to manage other humans. We've done that really well over the years. The shift that's happened over the past couple of months is that we now also have to train humans to manage agentic AI. When we complete the AI enablement for an agency and start looking at the opportunities that open up, like having an agentic AI call the phone, generate lead opportunities, or do live transfers, the question becomes who on the staff is ready to manage those agentic AIs. Right now, not many agencies have someone with that skill set in-house. To get in front of that gap, we've been hiring AI engineers at a frenetic rate. Think of them as fractional AI implementers who come into the agency to help with management, help with onboarding, and ideally get somebody on the agency's own team comfortable enough to take it over once we leave. Right now there's still a gap, because there aren't many humans inside agencies who are comfortable managing agentic AIs yet.”

If a retail agency owner is reading this and wants to get started with AI today, what are the concrete first steps you'd tell them to take?

“Let me start with what not to do. If you're going to use a public LLM, ChatGPT and Claude are the popular ones right now; don't try to use bubblegum and tape to connect it directly into your AMS platform. There's a reason AMS platforms aren't making it easy for tools like that to connect: the liability and risk exposure of putting private data into a public LLM is significant. So step one is making sure your data is truly yours as an agency, and making sure it's private and secured as you get AI-enabled. That's the question underneath the question of where to start: get your data AI-enabled, and make sure it's extremely secured while you do it. We've been working with a lot of agencies on exactly that. If an agency can figure that out on their own, that's great. If not, we're a resource. Right now at Agency VA, we're getting agencies AI-enabled so they can actually begin the game with AI.”

Scout InsurTech Thanks Our Partners