Case Study: Auto-Owners Insurance improves underwriting and claims processing with D2iQ

Auto-Owners Insurance Turns Unstructured Data Into Actionable Insights

Auto-Owners Insurance, a large mutual insurer serving policyholders through more than 44,000 licensed agents, needed a better way to analyze large, complex data sets for predictive modeling, underwriting, and claims processing. Its analytics team, led by Eric Truax, was running into limits with Hadoop and needed a faster, more flexible platform that could handle unstructured data such as photos, videos, and audio while supporting new tools like TensorFlow. They turned to D2iQ’s Mesosphere DC/OS Platform to support these data science and analytics workloads.

D2iQ implemented Mesosphere DC/OS Platform as the enterprise resource coordinator for Auto-Owners’ microservices and big data environment, running HDFS, Jupyter notebooks, and about 43 services across multiple business applications. The platform helped the team become more self-sufficient in DevOps, deploy models faster, and scale workloads on-premises while keeping the cluster around 80% capacity. As a result, Auto-Owners improved underwriting efficiency, reduced costs, accelerated its data science program, and made core business processes like underwriting and claims handling more effective through better use of unstructured data.

Auto-Owners Insurance

Eric Truax

Analytics