Chris Cheatham believes forming partnerships with QBE and Guy Carpenter were the two biggest milestones in the history of his company, RiskGenius, so far. But was there ever a time when he thought it was time to pack it in and give up the idea of an InsurTech connecting commercial insurance policy wording analysis and machine learning?

“One moment when I realized we were up against something much bigger than [we planned] was back in 2015, when we had launched the beta version of RiskGenius,” he recalled. “I knew very little about insurance at the time—just what I had gleaned from being a surety attorney. We had trained algorithms on these forms that a carrier had given us and then brought in users that were different. One new user started uploading documents—literally the first user started uploading documents—and I remember looking at the first document, and our algorithms had failed on it. ‘What is this?’ I thought. It looked so much different than what we had been looking at before.”

“We had trained on GL policies in the casualty space. The first live document that had been uploaded was a property policy. ‘Wait a second. We have to train on more than just GL policies?’ A perils clause? I don’t understand what this is.'”

“That was a big moment when we knew it was a much bigger undertaking than we’d realized.”

How did he find the courage to move forward?

“That question stresses me out even thinking about it,” Cheatham confessed. “We did what we always do, which is just evaluate the problem and figure out how you’re going to tackle it. Quickly, we had to figure out if there’s property policies, what else is there? You find there’s workers comp, and D&O, and specialty lines, and excess and umbrella, and commercial auto.” Armed with that knowledge, “you figure out what part of the problem you want to fix. And then you just start doing the work,” he said.

“I remember many, many evenings where I would stay up until 4 a.m., copying clauses and pasting them into spreadsheets, and typing out a category for that clause. You just keep doing it,” he said.

“At some point, you need to have a good, strong team that is willing to dig in on the problems,” he added. “But it’s not much more than just showing up every day and doing the work.”

Carrier Management also asked Cheatham to talk about the personal challenges that he might have faced in leaping from a past career of working in surety and construction law to tackle subjects he probably knew little about: machine learning technology, insurance. Coaxed to offer his advice for wannabe founders of insurance tech companies or non-tech insurance leaders who want to get into tech, he said, “No. 1: Find a problem that’s worth solving and that you can be passionate about solving.”

“If you find that and you can explain why you’re so passionate about solving the problem—which I was—then you can find people and experts that will help you solve the problem.”

“We have a chief data scientist who has helped me learn to understand machine learning—not that I’m a machine learning expert or data scientist by any nature. But at least I have a basic understanding of what’s going on. We found industry partners who took a lot of time to explain insurance to us and provided us a lot of information and documents and data.

“Without those people, this fails. You just need people that will take the time to help you understand.”

“Then the other thing is that one key habit of any entrepreneur: wanting to just constantly learn and understand things.” In his experience, “there’s more information out there than you could ever imagine about anything. If you want to learn machine learning, you can do it online with free resources at this point,” he said. “If you want to learn insurance, I guarantee you there’s more information than you can ever read and you can learn it very quickly.”

“You just keep grinding on it and doing it until you figure it out. But you have got to have a really strong network of people that are passionate about solving your problem, and then it becomes a lot easier.”

Cheatham sums up his leadership style with the phrase “Trust but verify,” noting that he picked that up from his dad, who ran a surety operation for a while. “I’m also really into sports. Andy Reid is the Chiefs head coach, and it feels like he’s hands-off if you do your job. I really like that tactic.”

“I try to always be trusting of people. If you go in not trusting, then immediately I think you have some cracks in the foundation…I go in expecting people to do their work. I will assume you’re doing the work until I see that you’re not. That’s when I’ll dive in and figure out what’s going on. The same with projects; the same with any initiative. I tend to be someone who is zoomed out and just making sure I understand what’s going on. But if I see an issue, then I zoom in and try to figure out what the problem is with you so we can address it and move on.”

These days, Cheatham is zoomed in on studying what can be inferred if you overlap economic data with insurance data and legal data. “What I’m thinking about a lot right now is [whether] there [is] a way to predict where COVID-19 claims and litigation are going to occur.”

“I don’t have an answer yet, but that’s what I’m studying,” he said. (CM is publishing a series of articles capturing what he learns in pursuit of the answer. The first in the series is “COVIDigation: Why So Many Cases? Are More Coming?“)

Tying this work back to the focus of RiskGenius, he said he recognizes that the answer isn’t necessarily tied to policy language, “but it definitely has implications on what you put in an insurance policy.”

Asked about what else might be on the drawing board for RiskGenius, specifically potential collaborations with other InsurTechs in coming years, Cheatham said there are some future partners that are intriguing.

“You’re seeing a lot of these storefront commercial insurance plays starting to take off, where people are willing to buy commercial insurance digitally. I think the next level for those types of companies is going to be real-time analysis of existing coverage vs. what’s being proposed by that company. What’s in my current insurance policy vs. what’s being offered?

“So, when I think about partnerships long term, those are interesting to me because we can understand the contents of policies,” he said. “Imagine somebody uploads a policy to the site and you can instantly say, ‘OK, my policy has these five gaps, and here’s how they compare to what’s being offered by this company that I just applied to.'”

In spite of the fact that Cheatham once wrote an attention-getting social media post about breaking up with InsurTech, he says he isn’t averse to the descriptor still being applied to RiskGenius. “It was [in] that time period where you were seeing ‘InsurTech’ being used for a lot of things—just legacy software that got slapped with that label of InsurTech,” he said, explaining part of the impetus for the temporary divorce.

Cheatham, who has also written articles about topics ranging from “How to Get 50,000 Followers on LinkedIn” to “What the development of cyber liability products can teach us about algorithmic liability,” said some COVID-19 commentary he offered in mid-March has drawn a lot of attention this year. “I came out fairly early and said this is going to be a big loss event. It was really early for saying that. And I’m not an actuary. I’d just done some basic math in my head.”