Additive manufacturing (AM) is the computer-aided design for the successive addition of layers by layer material. It is widely used because of the fast prototyping using laser metal deposition, which is difficult to implement using conventional techniques. Understanding the temperature profile prediction is necessary in AM processes, such as Bed Fusion process (PBF) technology to produce the right quality parts. Thus, the temperature profile prediction using Artificial intelligence techniques, like data-driven models and real-time iterative models using complex geometries, require real time control systems by considering the in-situ data. Besides, enhancing the accuracy of prediction is the hectic challenge faced by the existing systems. Hence, the proposed temperature profile prediction is developed based on an Artificial Intelligence method named Global herding algorithm-based neural network (global herding-based NN) to overcome the challenges associated with the existing methods. The proposed global herding optimization is developed by hybridizing the herding characteristics associated with the standard Elephant herding optimization (EHO) and Rhino Herd (RH) optimization to boost the solution's global optimal convergence. Moreover, the integration of the proposed global herding optimization with the NN model ensures the optimal selection of the hyper-parameters of the NN classifier, which renders effective performance of the temperature profile prediction. The effectiveness of the proposed model is revealed based on the performance metrics, such as MAE of 11.778, MAPE of 3.432, and MSE of 11.778.