Six years is a short lifespan for an insurance carrier. For Relm Insurance, it was enough time, however, to expose the limits of traditional operating models and force a rethink of how technology should sit at the center of the business.
“We have seen over 13–14,000 opportunities in more than 40 different countries around the world,” said Christian Davies (pictured), global head of distribution and innovation at Relm. “It was a scramble trying to build and build as the scale increased so rapidly.”
That pace of growth prompted a strategic pause late last year. Rather than continuing to add people and process, Relm’s leadership stepped back to assess how much of the business could be augmented, simplified or accelerated through artificial intelligence and other emerging technologies.
“The answer was almost everything,” Davies said.
Relm’s approach to innovation was deliberately pragmatic. Davies pushed back on the idea that insurance required a separate category of technology to modernize itself.
“I am not necessarily keen on the word ‘insurtech’, because I think there is a lot of native technology that could simply be brought into insurance as a 300-plus-year-old industry to assist with that,” he said.
From that perspective, AI and blockchain were treated less as experimental add-ons and more as practical operating tools. Having insured digital asset and emerging technology risks since inception, Relm already had exposure to the sectors driving much of that innovation.
“Bearing in mind we have been insuring this space for six years, there is absolutely no reason why we should not be doing it ourselves,” Davies said.
The first step was internal enablement. Relm rolled out a private, corporate version of OpenAI’s ChatGPT across the organization, giving all staff access to generative AI while keeping data contained.
“It was our own private version, where it did not lean on any data externally, and none of our data went out externally,” Davies explained.
From there, the company began building models in-house, focusing on underwriting efficiency and decision support.
One of Relm’s earliest deployments sat at the front of the underwriting process. Every submission was routed through a proprietary classification model built internally.
“We built our own GPT,” Davies said. “Every single one of our submissions ran through a model that classified what industry vertical it was, and whether it was better suited with our own internal appetite guide.”
That process also linked underwriting decisions back to regulatory and licensing requirements, an important consideration for a group regulated in Bermuda by the BMA.
For underwriters, the output was immediate. When a new account arrived, they were presented with what Relm called a “sector classifier assistant.”
“It gave them a two-page overview of what the sector was, how it aligned with our underwriting appetite, and whether there were any flags from a key underwriting criteria perspective that they could find online,” Davies said. “It summarised all of that.”
The intent was not to replace underwriter judgement, but to elevate it by reducing manual research and supporting faster, more consistent decisions.
“That was the first venture where we really went for it, and we did it ourselves,” Davies said.
Relm’s innovation strategy also accepted that not every initiative would succeed. While the company evaluated external providers, it prioritized internal understanding of the tools it used.
“We experimented with a lot of the products ourselves,” Davies said. “When we knew how to use them, we had a better understanding of how to underwrite them.”
That approach helped Relm move on quickly from ideas that failed operationally. One project with a third-party provider aimed to create an AI-powered internal knowledge repository.
“It turned out that the load time ended up being way too slow – around 10 minutes – so we decided we were not going to use that,” Davies explained.
Instead, Relm built its own internal “ask me anything” capability, embedded into existing systems.
“I could pull up information across our file structure, across Outlook, across Teams, and across our internal messaging or email system,” he said. “I could ask it a question and it would pull any relevant information from that in real time.”
The experience reinforced the importance of speed and usability over technical novelty.
What set Relm’s approach apart was how directly innovation was tied to leadership accountability. It was driven from the top, not delegated to a standalone team.
“It was coming from our CEO and founder, Joe,” Davies said. “It was embedded in the management framework.”
For 2026, every divisional leader was required to identify at least two areas of their business to augment with AI or other technology. Those commitments carried financial consequences.
“It was actually in every single one of the leadership KPIs for 2026,” Davies said. “People’s bonuses were tied to how they were utilizing and augmenting their day-to-day processes.”
That top-down mandate removed a common obstacle to transformation.
“In all honesty, we did not face any internal challenges,” Davies said. “There was no one in the company who said ‘no’.”
