Emergency department overcrowding is a patient safety crisis. Machine learning models that predict patient volume, length of stay, and bed demand hours in advance enable proactive capacity management and measurable improvements in throughput.