<p>Artificial neural networks are very useful for tackling nonlinear issues, which are common in fluid engineering. This makes them great for modeling nonlinear, dispersive wave motion, including how long waves, tidal bores, and internal waves interact in shallow coastal waters, estuaries, and continental shelves. This study investigates the nonlinear geophysical Korteweg-De Vries equation, a fundamental nonlinear wave model that has a wide range of applications in a variety of fields. The proposed model, which incorporates the Coriolis parameter, has applications in fluid dynamics for the description of shallow water waves and tsunamis, which are known as strong waves. These waves are generated by volcanic eruptions, landslides, or earthquakes that propagate across oceans. In order to acquire wave profiles for the studied nonlinear model, this investigation combines neural network approach with a modified generalized Riccati equation mapping technique. In this study, the neural network is a multi-layer computational approach that incorporates the activation and weight functions among neurons in the input, hidden, and output layers. Here, each neuron in the initial hidden layer is assigned the solutions of the Riccati equation. In this manner, the new trial functions are established. A variety of solutions in the forms of soliton profiles and other complex forms are secured by the application of the proposed technique. Moreover, a variety of graphs are presented to observe the impact of key parameters in this study. Changes in these physical and mathematical parameters influence the amplitude, shape, and stability of the obtained wave structures. The solutions obtained indicate the dynamic properties of the soliton deriving from the suggested model, which will aid further investigation into the associated phenomena.</p>