Abstract: As mining depths increase in underground operations, the risks of ground pressure hazards rise significantly, while traditional monitoring methods face challenges such as fragmented data and difficulties in integrating multi-source heterogeneous data. To address these issues, this study develops a ground pressure safety monitoring and early warning platform based on data processing analysis and 3D visualization technology. By integrating multi-source data including microseismic monitoring systems, stress sensors, surface displacement monitoring devices, and geomechanical databases, the platform establishes a three-dimensional dynamic correlation model between monitoring indicators and spatial locations. This enables real-time perception of ground pressure safety conditions and implements closed-loop management for tiered early warnings. During a 5-month operational trial at Maoping Mine, the platform successfully triggered 840 valid warnings, significantly enhancing the mine’s capability to prevent and control hazards such as rockbursts and roof collapses. The system has proven to be a reliable technical solution for ensuring safe and efficient deep mining operations.