Personalized medicine continues to gain momentum as a topic for discussion, yet directly linking patient-level decision support to more advanced analytical techniques, such as nonlinear mixed effects modeling, is not being practiced in most hospitals. Current practice for Vancomycin therapy uses dosing nomograms to determine the dosing regimen for patients. For simplicity, these nomograms stratify patients into bins based on some combination of weight, serum creatinine, and/or age to adjust starting regimens. Yet, studies across the US and Europe have shown as few as 37% of neonates achieve recommended target concentrations using such nomograms. The purpose of this research was to develop a bayesian decision support toolkit to provide adaptive, individualized dose recommendations for neonates. First, a bayesian nonlinear mixed effect model was developed and qualified for predictive forecasting in individual patients. Second, this model was used to develop a novel algorithm for dose individualization. Finally, a web application was developed to allow clinicians to provide decision support for clinicians involved in vancomycin dosing decisions. The proposed strategy can decrease the number of patients improperly dosed up to 90%, drastically reducing the chance for treatment failure, toxicity-related adverse events, and resistance development.