In this research, the removal of heavy metals from water sample from Ngomari Bus Stop in Maiduguri using low-cost agricultural waste biosorbent (watermelon rind – WMR) was assessed. The effect of contact time, adsorbent dosage and initial metal concentration against % removal and adsorption capacity were assessed. Atomic Absorption Spectrometry (AAS) analysis of the water sample reveals the presence of arsenic (As), cadmium (Cd), chromium (Cr), copper (Cu) and lead (Pb) at concentrations of 0.004, 0.170, 0.199, 0.286 and 0.201 g/100 mL before biosorption. Because an initial Fourier-Transform Infrared Spectroscopy (FTIR) characterization of the WMR exposes the presence of carboxylic and hydroxyl groups that can readily bound to metal ions, already proves it as an effective biosorbent for heavy metal removal from water. This leads to a considerable reduction in the concentrations of the above elements after biosorption, especially at an optimum biosorbent dose, water solution and contact time of 0.1 g, 100mg/L, and 120 min for As and 70 min other metals, respectively. The selection of the optimum conditions was facilitated by the use of Response Surface Methodology (RSM) Design of Experiments (DOE) existing data run for Gaussian Process Regression (GPR) analysis. The selected Radial Basis Function (RBF) combined with White Kernel during GPR analysis affirmed the effectiveness of WMR in adsorbing all the heavy metals at R2 > 0.995. A symbolic GPR model which is a second-order polynomial regression (quadratic) model for % removal and a first-order linear regression model for