This document discusses the development of an application called RxFx that provides personalized drug recommendations. It analyzes data from over 3 million self-reported entries on drugs, side effects, and indications from an open-source FDA database. The data is preprocessed to reduce the drug listings and calculate the probability of each side effect for each drug based on patient reports. The application then recommends drugs for a user based on their preferences by ranking side effects from least to most preferred and considering the probabilities of each side effect for each drug option.