The Emergency Department is perhaps the singular environment in a hospital where the greatest concentration of decisions are made with the least amount of clinical data; and the acuity-level assigned at triage can massively impact the trajectory of care for a patient. Additionally, Emergency Departments in the United States have experienced unprecedented levels of crowding over the last two decades, with correlation to worsened patient outcomes, preventable errors, and staff burnout. In 2017, the ED at Johns Hopkins deployed a CDS tool that leverages AI to generate risk-driven triage acuity recommendations, instead of relying on resource-based ESI. The tool has provided more reliable detection of patients with critical illness and reduced the time patients wait for care. They’ve also seen marked decrease in those assigned to the mid-acuity level 3, and instead more utilization of the low-acuity levels 4 and 5, which lends to better fast-track throughput for those patients who are to be discharged. Jeremiah Hinson, MD PhD, will speak of his experience developing and utilizing the tool with his team at Johns Hopkins.