Acute kidney injury (AKI) is a frequent adverse condition that increases patient risk for mortality. In 2013, approximately 47,000 Americans died from renal disease, which is becoming the 10th most common cause of mortality in the country. The purpose of this study is to improve the care provided to patients with AKI at Geisinger by leveraging data capabilities using top-of-the-line machine learning (ML) algorithms and patient interventions. Using outpatient and inpatient data, we will build models to assess patients’ risk of AKI at admission and severe AKI. New clinical pathways will be implemented using this information to improve outcomes of AKI patients.