The algorithm will be used in tandem with the company's access to ultra-high-resolution imagery. The technology was previously used by the Brit claims team and its delegated claims adjusters in the wake of Hurricane Ida. Brit has successfully deployed the technology to expedite the identification of insured property damage in the wake of the tornadoes that ripped through the Midwest Dec. 10-11. The machine-learning algorithm, developed by the company's data science team, assesses ultra-high-resolution aerial images and data. The algorithm allows Brit's claims team to identify, triage and assign response activity even before claims are reported.

Attorney involvement has been steadily driving up workers' compensation claims costs over the past decade, and research reveals the problem has never been worse. For example, in 2014, the California Workers' Compensation Institute (CWCI) published results of a five-year study that showed when a single injured party brought in a lawyer, the associated costs per claim went up by an average of $40,000 for permanent disability payments and $25,000 for temporary total disability benefits -- even if the case never went to court. In Florida from 2016-2017, legal fees related to workers' comp claims totaled nearly $440 million -- approximately $254 million of that came from employers defending claims while injured workers themselves were responsible for $186 million (an increase of 36 percent in just one year). And then there is the national scale. In a newly released study from CLARA analytics, data collected from a single national payer from injury years 2007-2017 was focused on closed indemnity claims across multiple states.

The technology can provide key improvements in workers' comp because it can detect new patterns out of things like natural text and images. Why should you care about machine learning (ML)? ML essentially refers to the phenomenon of computers and other devices that appear to be intelligent because they can learn from patterns in the environment to achieve a specified goal or predict an outcome. These patterns could come from a variety of sources, such as images, voice, free text or even structured data. The machine's ability to anticipate or interpret the information given can make it seem almost human-like.