Can software help insurance pros flag claims with nuclear verdict potential?

We hope you enjoy this excerpt from Litigation Daily, the exclusive source for sharp commentary on mega court battles, winning strategies and the issues that obsess elite litigators. Law.com subscribers can sign up for The AmLaw Litigation Daily newsletter here. Anyone else can click here to subscribe.

Back when we last chatted with Bob Tyson, the strategic managing partner of civil litigation and insurance defense firm Tyson & Mendes, he had just wrapped up the inaugural edition of the Nuclear Verdicts Defense Institute. That workshop for 18 defense lawyers and in-house counsel in the insurance industry drew on defense techniques outlined in Tyson’s book about fending off eight-digit jury awards.

Since that time, Tyson’s been busy. For one, he’s gearing up to host the second edition of the institute in June in the firm’s home city of San Diego — this time aiming to double the number of participants to about 40 lawyers. But Tyson has also helped launch a new venture in the world of insurtech. (With that pronunciation, think of insurance, not an insurgency, as I first did.) The company’s first product, which is set to launch shortly, is software aimed at helping insurance companies flag claims with potential high-dollar exposure should they proceed to litigation. We called up Tyson this week to discuss the new venture. 

What follows has been edited for length and clarity.

Lit Daily: So what have you been up to since we talked last July? 

Bob Tyson: We’ve created a whole new software company called Schaefer City Technologies. And our first product is about to go to market any day now. It’s a software product that can answer the number one question that I’ve received since my book came out: How can I spot a nuclear verdict in my claims file before it happens? So how can we spot a nuclear verdict as soon as we get the file and anytime up to and before trial? 

The more times I started getting asked this by insurance companies and claims professionals, the more I started to think about it. “Hey, is this something that we could figure out? Is it formulaic?” And as it ends up it is. We were able to use our expertise and our research on nuclear verdicts and then we partnered with Villanova University and Nathan Coates, who’s a professor of business analytics. With him and his students and his research body, we’ve been able to create an algorithm that can detect nuclear verdicts before they happen. So it’s like an early warning/early detection system for insurance companies that can attach to their existing claims systems to help them to be warned about cases that could go nuclear. So when you combine the business analytics and the research with our nuclear verdict research and our trial experience from trial lawyers that are in the trenches trying these cases, that combination has been a software product called NaVel. That is the first product of Schaefer City Technologies that can spot nuclear verdicts before they happen.

Where’s the name Schaefer City Technologies come from?

Great question. So, I grew up in Staten Island, New York. And when I was growing up, my dad liked Schaefer Beer. Schaefer Beer was a low-end …

Oh, I’m aware. 

I’m sure. There were these commercials where as long as you had Schaefer Beer wherever you were, you were in the Schaefer City state of mind — like you’d made it. 

So there’s one commercial that jumps out to me where there were two guys in beach chairs and they had a beach umbrella and they pop open their Schaefer Beer. And you think they’re somewhere in Maui or something. But as you pull back, they’re on the top of an apartment building in Brooklyn. But what they’re saying is, “Hey, you made it, you’re in Schaefer City.” So it’s kind of a play off that. We’ve created something of value. We’re there. You’re in Schaefer City now. You’ll be good. So not everyone’s gonna get the reference. But I’m in this business with my law partner, Pat Mendes, my law partner, Cayce Lynch, the professor from Villanova and a tech [professional], and then my sister Denise Tyson — she and I both grew up and are familiar with the Schaefer City references — is the CEO.

So what is the data set that you had the professor look at? 

What the professor looked at was more than 100,000 closed files, either settled or through jury verdict. We sourced that from one of the top three insurance databases in the world. We had to purchase all the data of course. 

It is machine learning. It is AI. So, the more data we get — we had over 10 million data points that were going through this machine learning process — the better the program gets. Right now it’s ready to go and we’re going to market this week, I believe.

So, what does the tech do? Do you know from looking at what NaVel puts out what could spark a nuclear verdict? Or is it something where only the algorithm knows to flag it and give you a rating?

Good question. Part of the software program is that every claim file that is run through the system gets a score from one to 10. And that score then is color-coded. The lower end of that is green, and you don’t really have to worry about that when it comes to nuclear verdicts. Once you start to get into yellow and red, that’s when management needs to take a look at these files and address why the file is at risk. What we share with them is the three main reasons that any particular file may have triggered a higher level score. We do share with them some of the data points that point to why this claim is at risk of going nuclear.

So you’ve got something in its nascent stages that’s flagged as something that has the hallmarks of a case that could go nuclear. What next?

So this is meant to be a tool to assist insurance companies and their claims departments in detecting cases that might be at risk for going nuclear. So the hope is that by using this tool, you will catch files very early on in the process that can get the attention of seasoned claims professionals, get the attention of management, so that if it needs to be addressed, if you need to resolve it sooner than later, if it’s a case that should never go to trial, you’ll be able to figure that out very early in the claims cycle. That’s the goal of this algorithm: to be another set of eyes for insurance companies to find these problematic files that could potentially slip through the cracks. That’s what you don’t want to have happen.

So what does it plug into on the insurer side? 

This software attaches to the insurance company’s existing claim system. We don’t store any of the data. We don’t house any of the data. The data is the insurance company’s data. They keep that. There’s no need to input any additional data. So it’s not like the claims people have to type in information so that NaVel works. They don’t have to do that. NaVel works in the background. It’s not creating any additional work. It’s just meant to be an additional resource and tool to help the frontline adjuster who will be able to see on his screen: “Oh, here’s the score for this claim. It’s a three or it’s a four.” 

Then it’s also meant to be a tool for management because there’s a management portal where they can see what the scores are for all their claims, not just one claim that’s open on their desk. They get to see the score for every single file that’s already on their system. And then they can use that data to see how many potentially risky claims are being handled by one team or one adjuster and move around resources internally and make sure that supervisors are taking a look at this case on this team or this other case on that team. So it’s an aid to an adjuster on a particular file. And we also have a management portal. Where they can see all files.

So you’re dealing with these things at the insurance claim stage, rather than the litigation claim phase, right?

Our system starts as soon as you open a new file. And when you open a new file, you don’t have as much data as you do once it’s in litigation. So what happens is the score can change over time. So every time more data comes in, NaVel runs in the background and gives a new score. So maybe it becomes riskier as time goes on and the score gets higher. Or maybe it becomes less risky as more information comes in. But it starts right at the inception of the claim, well before litigation. But then it also works throughout litigation too. So for the life of the claim from the beginning, when it might just be a police report or first notice of loss all the way through up to and close to trial.

Related:

Notes from ALM’s Complex Claims & Litigation Forum 2023

The problem with third-party litigation financing

Pain and suffering: 10 of the toughest verdicts to hit the insurance industry

Runaway verdicts: Reversing the trend