This research investigates the mechanical characteristics of epoxy matrix composites augmented with graphite filler particles at different loadings (0–7 wt.%). Results indicate that tensile and flexural strength significantly increase at 3 wt.% graphite loading, followed by a decline at higher concentrations. Similarly, impact resistance and hardness improved up to 3 wt.%, with no further enhancement beyond this threshold. Artificial Neural Networks (ANN) were employed to develop predictive models, demonstrating high accuracy (R > 0.96) across the training, testing, and validation phases. The ANN model effectively forecasts the mechanical behavior of graphite/epoxy composites, aiding material design optimization. These findings highlight the optimal graphite content for enhanced performance and the potential of machine learning in material science applications. This research advances the development of high-performance composite materials with customized mechanical properties for diverse engineering applications.