 DeepServe is a Cox proportional hazards deep neural network that models interactions between a patient's covariates, clinical and genetic features, and treatment effectiveness to provide personalized treatment recommendations. It has been shown to perform as well as or better than other state-of-the-art survival models in simulated and real survival data, and can model increasingly complex relationships between a patient's features and their risk of failure. DeepServe can also be used to model the relationship between a patient's features and effectiveness of different treatment options, enabling personalized treatment recommendations that increase survival time for patients. This article was authored by Jared L. Katzman, Yui Shaham, Alexander Kloninger, and others.