Customer Stories

How Auto-Owners Insurance Turns Big Data Into Big Opportunities [Case study]

Big data and machine learning drive insurance modeling for lower risk and faster claims processing

Nov 19, 2018

Alex Hisaka


5 min read

We have chosen to sunset DC/OS, with an end-of-life date of October 31, 2021. With D2iQ Kubernetes Platform (DKP), our customers get the same benefits provided by DC/OS and more, as well as access to the most impressive pace of innovation the technology world has ever seen.

Learn more about D2iQ Kubernetes Platform here.
Big data and analytics is what makes Auto-Owners Insurance possible. Without the ability to analyze large, unstructured data sets about customers, insurance companies like Auto-Owners Insurance are unable to calculate the risk profiles on which their business depends.
The problem is, though, that the potential in terms of risk assessment of data collection and analysis is an enormous one. Many insurance companies don't have a scalable way to incorporate many types of data into their prediction models. This can impact the ability to properly assess and minimize their risks, but also tailor more insurance products that suit each customer's needs and expectations.
"The business of insurance is about keeping a promise to our customers to be there when they need us," said Eric Truax, Analytics Manager at Auto-Owners Insurance. "We're always looking to find new ways to analyze data and use predictive modeling to improve our customer service, underwrite policies more effectively, settle claims faster, and better match premium to risk."
Eric Truax, and the Auto-Owners analytics teams, were looking for a way to analyze larger, more complicated data sets than their current systems could handle. The data sets that they wanted to analyze often involved unstructured data, like photos, videos, audio files, and more. Truax and team turned to Mesosphere DC/OS, which offered the kind of support that a large enterprise, like Auto-Owners, required. As the team kept expanding its cluster and discovering new uses for their work around the company, the agility of DC/OS in terms of operations, development, and data began to shine through.
"We keep throwing more and more problems at DC/OS and getting more value out of it. It's a great fit for us," said Truax.
For Truax, a big advantage to using their DC/OS cluster for predictive modeling is that the development group can make and deploy models to production and do some of the IT work themselves, cutting down on time to market and decreasing their reliance on other teams. The team now considers itself a self-sufficient DevOps team rather than just a development team, which has turned out to be a big value proposition.
"After an initial learning curve, everyone came to realize we knew what we're doing, and we didn't break anything important," said Truax. "Over the past year, we've onboarded my entire team to our cluster which keeps the system pegged at about 80 percent capacity.  The occasional bursts have underscored our need to expand the cluster resources again this fall."
With new Insurtech disruptors popping up to improve various aspects of the insurance process, established insurers, such as Auto-Owners, must respond to this new competition as soon as it appears. Thanks to its DC/OS implementation, Auto-Owners is able to use the insights gathered from analyzing unstructured data to improve its underwriting efficiency, reduce costs, and accelerate its data science program.
"Insurance has always been data-driven, especially for pricing," said Truax. "Today, we are using DC/OS to make many of our core business processes like underwriting and claims handling more effective by leveraging the data we've captured and becoming more data-driven than ever."
To learn more about Auto-Owners and how they leveraged DC/OS, download the case study.

Ready to get started?