Injecting water into the ground to produce geothermal energy can trigger small earthquakes, which raises safety concerns. This study focuses on a project in Utah to better predict when faults—natural cracks in the Earth—might slip and cause these events. We group earthquake signals to identify the shape and direction of underground faults. Then, we assess how likely those faults are to slip by using both traditional physics-based models and a machine learning model trained on millions of examples. This approach helps estimate fault behavior even when laboratory measurements are not available. We apply these methods to recent data from the Utah site. Our results show that we can map where faults might slip and use this information to manage risk in real time, helping make geothermal energy development safer.