COVID-19 has over only a few short months resulted in the deaths of >400,000 people worldwide (as of mid-June 2020) and has forced many others to suspend their livelihoods. A vaccine is unlikely to arrive until sometime next year, but a safe and effective drug could be developed, manufactured, and distributed before that. With this project, we are working to develop a therapeutic for COVID-19 that is safe, effective, and difficult for the SARS-CoV-2 virus to evolve resistance against. By taking the receptor that the virus normally uses to enter our body's cells (ACE2) and tweaking it so it can be made into a drug—making a so-called "decoy receptor mimic"—we are diverting the virus away from the normal ACE2 receptor, leaving it unable to replicate itself in cells. To do this, we are using a combination of laboratory evolution and machine learning techniques to rapidly identify hundreds of candidates with high binding affinity to the viral spike protein and—in order to ensure safety—low predicted binding affinity with other proteins in the human body. We are hoping that this strategy will provide a blueprint that can be extended to proactively develop decoy receptors for other viruses, enabling us to get ahead of the next pandemic.