“Most algorithms stop once they make the prediction,” she said. “But we take it a step further than anyone else by looking at the practical implications of what it means to monitor a patient.”