<p>Because of its exceptional mechanical strength, corrosion resistance, and thermal stability, the nickel-based superalloy Inconel 718 is widely utilized in the aerospace, marine, and energy industries. However, attaining ideal surface integrity is difficult due to its poor machinability, which is brought on by low thermal conductivity, severe work hardening, and strong tool adhesion. In this work, data-driven methods are used to examine and forecast process outcomes during die-sinking electrical discharge machining (EDM) of Inconel 718. The machining of Inconel 718 is performed to a depth of 0.5&#xa0;mm using a copper tool electrode in submerged kerosene dielectric fluid. Material removal rate (MRR), tool wear rate (TWR), and surface roughness (<i>R</i>_<i>a</i>) were taken into consideration as responses, while pulse-on time (<i>T</i>_on), pulse-off time (<i>T</i>_off), and peak current (<i>I</i>_p) were chosen as input factors. Predictive modelling was done using response surface methodology (RSM) with Minitab 17 and radial basis function neural network (RBFNN) using MATLAB 2020b. From the Analysis of variance (ANOVA), <i>T</i>_on, <i>T</i>_off, and <i>I</i>_p are all observed as influential for MRR, TWR, and <i>R</i>_<i>a</i>. The values of the coefficient of determination (<i>R</i>²) for MRR, <i>R</i>_<i>a</i>, and TWR, respectively, predicted by the RSM technique are 0.93, 0.96, and 0.98. The same predicted by RBFNN is 0.94, 0.99, and 0.97. This indicates that the implication of data-driven techniques is better for predicting the non-linear behaviour of the die-sinking EDM machining. Field emission scanning electron microscope (FESEM) observation shows the formation of recast layer, micro-cracks, and micro-holes on the machining surfaces. The micro-hardness of the machined specimen increases by 200% more as compared to before machining. Energy dispersive spectroscopy (EDS) analysis shows that the content of iron and chromium increases, which indicates a better hole strength after machining.</p>