Case progression for individual patients is exceedingly complex. Identifying at risk patients early allows mitigation sooner and better health outcomes while avoiding needless procedures. While most of the information needed to identify these patients will be stored in the Electronic Medical Record (EMR), this data may be stored in different tables at different granularities. The challenge here is to bring all this data together, align it to a single patient at a moment in time to allow the machine learning algorithms to build the prognosis model. This data will include patient demographics and information from admission, patient history, patient vital signs over time, procedures performed, labs ordered and results, and doctor notes. All of this information will need to be cleaned, joined and transformed into valuable ML features before going into model training. This pre-modeling prep process can be frustrating and time consuming. We are here to help.