When associated with certain biological conditions, co-variation in concentrations for trace molecules present in the exhaled human breath provides discriminating health information that can be used for medical diagnostics. We have developed a state-of-the-art breathalyzer based on an optical frequency comb laser, capable of broadband spectroscopy data collection at rotational-vibrational quantum state resolution and at parts-per-trillion (10^-12) record-level detection sensitivity. We have carried out the first field research study over a total of 170 individuals with near 50:50 class distributions for COVID-19 (positive/negative) and found an excellent prediction power obtained through measurements of breath. In this talk, we will discuss how we have brought optical frequency combs into medical science. We will cover the principle of comb spectroscopy with cavity enhancement, supervised learning data analysis, and exciting new opportunities for communities in laser science, molecular spectroscopy, and computer science. This talk will be organized to be accessible to audiences from vastly different research disciplines.