Data Scientist
CurrentDeveloped and deployed models to precisely predict patient responsibility amounts for XiFin customers with 98% accuracy.Leveraged NLP, OCR, and machine learning to process insurance card images in order to streamline insurance claims process for end users.Took ownership of a monthly data load process, automating the load and validation of terabytes of Payor claims data for use in business dashboards.Productionized a machine learning model to accurately assign error codes to the correct user for resolution.Designed a NLP model that matches status claim codes to the correct response code.Spearheaded anomaly detection efforts for productionized models to identify model drift and abnormal data points, retraining models when necessary, ultimately improving previous accuracy by up to 5%.Work closely with business stakeholders across various departments to pinpoint potential machine learning solutions to increase business efficiency.Used NLP and OCR technologies to scan and classify 90,000 documents for internal indexing
