ClaimScore

ClaimScore is the only independent software solution dedicated to resolving the ever expanding claim fraud problem in Class Action Settlements. We have combined a 65+ point expert-system artificial intelligence algorithm with a neural-network machine learning system to ensure the highest level of accuracy possible; literally beyond humanly possible. This accuracy is stacked on a complex cloud-architecture that can live review 1,000's of claims/second, indefinitely. ClaimScore's Proprietary algorithm means each claim receives the same consideration. Each claim begins with a ClaimScore of 1,000 and is reduced each time it fails a criterion. Each criterion has either a fixed weight or sliding weight depending on both the correlation to fraudulent claims and correlation to valid claim. Once a claim is scored, it receives a Recommended Determination. If a claim’s ClaimScore drops below 700, ClaimScore recommends that the claim be rejected as non-compliant with the specific terms of a Settlement Agreement, including an analysis that the claim contains “indicia of fraud”. A 700-passing score was selected to allow claims to fail certain criteria, yet still be approved; furthermore, claims are not rejected for single criterion unless they are specified in the Settlement Agreement. Claims are all tagged with the relevant Deduction Codes. To maximize transparency, each claim is tagged with deduction codes associated with the criteria it fails, thus ensuring that the parties, the administrator and the Court definitively know all specific reasons why each claim was rejected.