The global demand for sustainable energy drives the need for alternative fuels, with biodiesel emerging as a promising candidate because it is renewable and eco-friendly. In this study, an optimized biodiesel production process was developed using a blend of Palm kernel oil (PKO) and Jatropha curcas oil (JcO), catalyzed by KOH-supported calcined animal bone waste (KOH/CABW). A response surface methodology (RSM) technique, based on a rotatable central composite design (RCCD), optimizes the transesterification reaction. The variables studied include a methanol-oil molar ratio (v/v), catalyst load (wt%), reaction temperature ( °C), and reaction time (min), with biodiesel yield (%) as the response variable. A Genetic Algorithm-Bayesian optimization (GA-BO) hybrid approach is employed to further enhance biodiesel yield. Fuel properties of biodiesel and catalyst reusability studies are conducted. The result from the RSM analysis, supported by ANOVA, reveals significant statistical relevance of the quadratic model at a 95 % confidence level, accounting for individual process variables, and interactive and quadrative effects. The optimal biodiesel yield from RSM is 86.76 % at optimized conditions. In comparison, the GA-BO hybrid approach results in a higher biodiesel yield of 96.45 %, at modified conditions. Experimental validation of the GA-BO approach further confirms a biodiesel yield of 96.67 %, with fuel properties meeting international biofuel standards. Catalyst reusability studies demonstrate that the KOH/CABW catalyst remains effective and efficient after several transesterification cycles. The findings in this study present an innovative approach to biodiesel production by blending non-edible oils, utilizing advanced optimization techniques, and offering a sustainable energy alternative with minimized environmental impact.