Earthquakes are among the most destructive forces of nature. Their occurrence leads to widespread destruction and loss of life, making them a critical natural hazard globally. While seismology has advanced, accurately forecasting earthquakes remains a major challenge due to the complex external and internal factors involved. Rather than short-term prediction, long-term earthquake forecasting has emerged as a vital research direction. With the advancement of artificial intelligence and machine learning, data-driven models can now be employed to address this challenge. This paper proposes Quake Predict, a machine learning-based framework that uses historical earthquake data to forecast seismic risk zones. The system applies Random Forest Classifier for spatial risk classification and identifies high-impact regions. The study also integrates feature selection techniques and performance evaluation metrics to ensure accuracy and reliability.