In this paper, the systematic computational design process of AlGaN-based multiple quantum-well (QW) deep-ultraviolet (DUV) light-emitting diode (LED) grown on sapphire (Al2O3) substrate was investigated. An optimization was held to increase internal quantum efficiency (IQE) handling the LED parameters such as doping percentage of the n- and the p-type layers of these devices. The structure parameters of the best design were determined through a customized genetic algorithm integrated into the nanostructure quantum electronic simulation (nextnano). As a determining factor, IQE was obtained to be 24% for the devised 285 nm LED. It has been demonstrated that this result can be increased up to a remarkably high value of 70% by a low threading dislocation density (TDD) and reduced Auger recombination. In addition, the operation input power and potential difference were successfully kept below 0.1 W/mm2 and 5.05 V, respectively.