Machine learning can automate pieces of the claims denial process that are often costly to maintain with a traditional model. Most benefits management companies rely on groups of doctors and nurses to help identify fraudulent claims and approve or deny incoming requests. The goal of Enterprise AI is to magnify these efforts using advanced machine learning techniques and push only the most relevant and complex claims to human experts in the process.
eviCore Healthcare was able to utilize the Chorus Platform to automate several steps of the claims approval process using machine learning. In the past, operationalizing models alone would take a few months. Now, their business analysts are able to build models of their own using the Visual Workflow Editor and push them into production without needing to engage in a lengthy development process with IT. This use case alone helped eviCore achieve 500% process improvement.
“I worked with eviCore to put a classification model at the very start of their process so that routine claims are immediately approved when the model has high confidence for approval. Now we’re building new models that use unstructured data, like nurses’ notes, so that AI can help out in deeper process stages where there has already been some back-and-forth communication with the provider.”