A BUILT Healthcare client wanted to gain more visibility on the financial impacts of ongoing Payment Integrity and Fraud, Waste and Abuse issues amongst its members. During the COVID crisis with the increased demand on healthcare there was a recognized need to identify potential problems as early as possible.
BUILT developed models in their Azure Data Lake and Data Bricks layers utilizing R, Python, SQL and PowerBI to profile the data, score it, and develop reliable reports and dashboards to perform advanced statistical analysis and modeling utilizing linear and non-linear regression models as well as resampling and Markov chains.
Our team was able to build predictive models using machine learning to identify high risk transactions for further triage and investigation. These dashboards are now widely used within the organization to halt suspicious activity early in the process. This saves the organization millions each year by preventing “after the fact” triage, re-stating and re-processing claims data multiple times as well as the $$ lost from the fraud that previously would have gone un-noticed.
Photo by Anna Nekrashevich on Pexels