Insurance and risk veteran Jim Lewis, Executive Director of the Vaughan Institute at The University of Iowa, breaks down why the biggest cost in a self-insured employer’s workers’ comp structure is the one nobody is trying to prevent.
The Financial Blind Spot in Self-Insured Workers’ Compensation
For self-insured employers, the single largest cost in their workers’ compensation structure is the claim itself. Administration, support, and the infrastructure to service claims over time account for roughly 30% of total cost. The rest is claims, and the majority sit in the retained layer, fully absorbed by the employer.
On this episode of the Frontline Advantage, host Vivek Kumar sits down with Jim Lewis, Executive Director of the Vaughan Institute of Risk Management and Insurance at The University of Iowa’s Tippie College of Business. The two first met in 2019 through the gener8tor Minneapolis accelerator and reconnected in London, where Jim got a firsthand look at Teamforce AI’s latest capabilities.
A “Crowded” Space, and What’s Missing From It
Jim draws a clear distinction between the current landscape of insurance technology solutions, which he calls “crowded” with tools focused on efficiency, speed, and customer experience, and the far less crowded territory of actually predicting and preventing claims before they happen.
As Jim puts it: “When you get at actually predicting and preventing, ooh, that is very interesting. And you have the data from what I’ve seen that can do that.”
He describes most of the industry as sitting on reactive claims data that makes it nearly impossible to predict or prevent outcomes. The missing piece, in his view, is signal data from the frontline workforce that surfaces risks and blind spots before they become recordable incidents.
The 10% That Changes Everything
Jim puts a specific number on the opportunity. If a self-insured employer could prevent just 10% of claims from occurring, the financial impact is enormous, particularly when those claims sit entirely within the retained layer below the stop-loss threshold. That reduction flows directly to EBITDA, not as a safety initiative, but as a financial event.
He notes that for risk managers at self-funded corporations, especially those with exposure in manufacturing, logistics, and factory environments, this kind of data represents the single most important signal available.
Empowering the Individual as Risk Manager
One of the most striking points Jim makes is around what he calls “collaborative innovation:” empowering frontline workers to become active participants in risk management rather than passive subjects of it. He describes this as something the industry has historically found very difficult to achieve and identifies Teamforce AI’s approach as a way to finally make it work at scale.
Jim compares this to other insurtech approaches using wearables and cameras but notes the simplicity and directness of collecting signals from the workforce itself. His reaction: “It’s just kind of one of these like, almost like, why didn’t anyone think of this before?”
Cameras, Captives, and the Legal Question
The conversation also covers significant ground on adjacent topics. Jim discusses IntelliSee, a company using AI and existing security cameras to identify slip-and-fall patterns, noting that carriers are already offering discounts for camera-based prevention tools. But he and Vivek explore a critical limitation: cameras can capture a hazard, but they cannot verify whether the fix was implemented and sustained. That verification gap is where the proof loop matters.
Jim identifies captive insurance programs as a particularly strong market for prevention-focused technology, citing his recent conversations with Captive Resources in Chicago. Captive members are collectively self-insured, aligned on risk reduction, and typically more innovation-friendly than traditionally insured companies.
The episode also addresses the legal dimension of undetected risk. Vivek raises the issue of “plausible deniability,” where some companies avoid surfacing data for fear that knowing about a risk and not acting creates liability. Jim reframes this sharply, arguing that employers should focus on the revenue side of prevention rather than worrying about the small percentage of cases where data could be used against them in court.
Why Workers’ Comp Is Becoming a CFO Priority
Vivek shares that CFOs who previously would not have taken his call are now proactively engaging when the conversation is framed around retained loss reduction rather than safety or engagement. Workers’ comp costs are rising even in environments where safety incident rates are flat, driven by healthcare cost inflation. For low-margin industrial businesses, reducing retained losses has become one of the most direct paths to improving the P&L without requiring revenue growth.
Jim validates this shift, noting that when the conversation moves from selling safety in the OpEx budget to selling EBITDA protection, the entire dynamic changes. He describes the opportunity as “relatively low risk, high impact, and not much downside.”
From Classroom to Factory Floor
The episode closes with Jim offering to bring Teamforce AI into his insurtech innovation course at the University of Iowa as a student project, noting that his undergraduates could likely validate the ROI model with the data Teamforce AI is already producing. The Vaughan Institute’s program has grown from 50 students to over 500 under Jim’s leadership and is now the third-largest risk management and insurance undergraduate program in the country according to Business Insurance magazine’s 2025 survey.
Jim’s final assessment: “I think this is a winner that you’ve been working on for a while.”
Watch the Full Episode
Dive deeper into Jim Lewis’s analysis of the self-insured workers’ comp opportunity and his perspective on why frontline signal data is the most important dataset the insurance industry doesn’t have yet.
Connect with Jim Lewis on LinkedIn.