Abstract Wire and arc additive manufacturing (WAAM) is a predominant technique for metal based additive manufacturing process. In this process, generation of layers by depositing beads puts high emphasis on single bead geometry. This paper focuses controlling the bead geometry by suitable selection of welding process parameters. For this artificial neural network (ANN) is modeled to predict the bead parameters based on given process parameters and then the reverse model is designed to select the weld parameters based on user specified bead geometry. The results show that the variable welding parameters are significant and the ANN model assures the possibility of predicting welding process parameters for desired bead geometry in WAAM application