Pulsating flame jets have been widely used in open-top carbon anode baking furnaces for aluminum electrolysis. Reducing energy consumption and pollutant emissions are still major challenges in baking (heat-treatment) carbon anode blocks. It is also of immense significance to bake all the anodes uniformly irrespective of their position in the furnace. Baking homogeneity can be enhanced noticeably by optimizing anode baking operational, geometrical, and physical parameters. In the present study, CFD simulations are combined with a response surface methodology to investigate and optimize the effects of pulse pressure, pulse frequency, and mainstream inlet oxygen concentration and mainstream inlet temperature. Two-levels half fractional factorial design with a center point is employed. It is perceived that pulse combustion with short pulse time and high momentum results in significant enhancement of the anode baking furnace energy efficiency. The temperature homogeneity is also significantly improved. It is found that the oxygen concentration is statistically the most significant parameter on NOx and soot formations, followed by the fuel flow rate. For NOx formation, air inlet oxygen concentration has a strong interaction with pulse duration. Coupling CFD models with the response surface methodologies demonstrated great potential in multi-objective optimization of the anode baking process with enhanced energy efficiency and baking uniformity.