In this video, see how we helped a top five US Property & Casualty insurer uncover hidden subrogation opportunities in challenging PIP claims. By leveraging advanced generative AI models, we identified a critical alert that enabled the insurer to recover payments that would typically be overlooked due to stringent legal criteria. Our innovative approach not only streamlined the evaluation process but also ensured accurate identification of fault, ultimately saving valuable time and enhancing recovery opportunities. Watch the video to see how our technology transforms the claims process and empowers insurers to maximize their recovery potential!
This video is part of a series of interesting cases, featuring several of our insurance-focused team of data scientists.
If you don't see the video above click the button below to adjust your cookie settings
Manage Cookies"So recently we were working for a top five US P&C insurer and we identified a very interesting alert in the state of New York related to subrogation. What was remarkable about this particular alert was that it was actually on a personal injury protection or PIP claim which in the state of New York, for background, is very difficult to recover. The reason for this is that legally in the state you are only allowed to recover on PIP payments if a few very complicated criteria are met.
Namely, one of the involved vehicles has to either be a commercial livery vehicle or exceed 6,500 lbs. in weight. This is difficult to verify at scale. What we were able to do at Shift for this particular claim was to come in and take the available third party vehicle information associated with the claim and compare it against some external data sources containing registered commercial and livery vehicles in the state of New York. In doing so, we automatically identified the third party vehicle as a registered ride share vehicle, thus making the claim eligible for subrogation. Once this was identified, we took the claim, ingested it into our standard detection pipeline, which contains generative AI models, and those generative AI models looked at the notes associated with the claim, and determined that the third party ride share vehicle was in fact at fault for the accident.
Thus, all of the criteria necessary for a subrogation recovery were met and we were able to deliver an alert to the insurer that they in all likelihood would have overlooked. So for this particular case, what generative AI was able to do was evaluate the circumstances of the claim both quickly and accurately. If the insurer was evaluating this claim manually, they might have had to read through all of the claim notes and text associated with the claim in order to make heads or tails of who was at fault in the accident. So from that angle, generative AI saved the insurer a lot of time reviewing that claim.
Alternatively, if the insurer was using text analysis techniques that were less state-of-the-art than the generative AI that we currently use, they would have run the risk of potentially missing the subrogation opportunity that we identified, or they would have evaluated the claim incorrectly and also wasted some of the time of their own reviewer's time."
