Tungsten inert gas (TIG) welding is a highly significant arc welding process that utilizes a non-consumable tungsten electrode and an inert gas for arc shielding, resulting in excellent weld quality and precise welding operations. In structural applications, dissimilar metal joints, such as those between stainless steel and mild steel, are widely used due to their favorable mechanical properties. However, joining these materials presents challenges related to differences in composition, thermal properties, mechanical properties, and the formation of intermetallic compounds (IMCs) at the joint interface. This study focused on optimizing TIG welding parameters (i.e. welding current, arc voltage, welding speed, and gas flow rate) for joining AISI 1020 mild steel and 304 stainless steel 2mm thickness sheet metals. The developed automated TIG welding fixture improves weld quality by maintaining constant welding speed, minimizing defects associated with fluctuations in human-operated torch movement. The automated TIG welding fixture can be used to weld sheet metal in a single pass by linear motion. Three levels of variation were applied to the selected input parameters, and nine experiments were carried out using Taguchi's L9 orthogonal array approach. An Artificial Neural Network (ANN) was constructed to model the automated sheet metal TIG welding process, with a 4-10-3 network trained by the Levenberg-Marquardt algorithm mean squared error (MSE) of 1.0069e-1. The objective was to increase the hardness, tensile strength, and bending strength of the weld quality. The Multi-Objective Genetic algorithm (MOGA) was used to determine the combination of optimal process parameters, which was identified as welding current 91.835 A, welding voltage 13.856 V, welding speed 19.850 cm/min, and gas flow rate 5.388 L/min resulting in maximum achieved values of Ultimate tensile strength 539.516 MPa, Rockwell hardness 92.449 HRB, and bending strength 731.216 MPa. Finally, a confirmation test was conducted with the optimum parameters. The predicted and confirmation test results percentage error was 0.23 % for tensile strength, 0.82 % for hardness, and 0.18% for bending. Thus, it has been concluded that the confirmation experimental results are within the acceptable range of percentage error as per the reviewed literature. Keywords: Dissimilar metal joints,TIG Welding process parameters, Artificial Neural Network (ANN), Multi-Objective Genetic Algorithm (MOGA), mechanical properties.