Cardiovascular (CV) manifestations of COVID-19 infection carry significant morbidity and mortality. Current risk prediction for CV complications in COVID-19 is limited and existing approaches fail to account for the dynamic course of the disease. Here, we develop and validate the COVID-HEART predictor, a novel continuously-updating risk prediction technology to forecast CV complications in hospitalized patients with COVID-19. The risk predictor is trained and tested with retrospective registry data from 2178 patients to predict two outcomes: cardiac arrest and imaging-confirmed thromboembolic events. In validating the model, we show that it predicts cardiac arrest with a median early warning time of 14 hours and an AUROC of 0.93, and thromboembolic events with a median early warning time of 168 hours and an AUROC of 0.85. The COVID-HEART predictor is anticipated to provide tangible clinical decision support in triaging patients and optimizing resource utilization, with its clinical utility potentially extending well beyond COVID-19.