The present research work deals with the influence of totals of comprehensive stresses, steady-state pore pressures, effective stresses, and depth of excavations on pore water pressures during construction, specific to deep excavations. This makes pore water pressure a great concern in the stability of the soil; hence, an important controlling factor in infrastructure projects. The hybrid technique being derived in this study involves the combination of Finite Element Analysis (FEA) and Artificial Neural Networks (ANN) modeling. First, FEA is done in which the mechanical factors distribution detail data in the soil under different conditions is simulated, while afterward, the ANN model will be sought for training on the results to provide a better prediction concerning the variations in the pore water pressures, taking the input from FEA that reduces the computational load and enhances the precision in the predictions. The results of the analysis showed that total stress and depth of excavation are the most dominating variables that govern changes in pore water pressures, while effective stress is also a significant factor; it shows relatively less impact. The ANN model showed high predictive performance with an R-squared score of 1, thus maintaining an absolute prediction accuracy for the simulated conditions. This research is very important in deep excavation construction projects, providing a powerful tool for risk prediction and control. The application of ANN to FEA data opens new avenues toward the optimization of the pore water pressure prediction process, allowing increased safety and efficiency in the execution of the most complex geotechnical projects.