Rotary Friction Welding (RFW) is essential in manufacturing automotive and marine components, yet the low-cycle fatigue life of dissimilar carbon steel joints (C35 and C45) remains underexplored. This study investigates the influence of RFW parameters on fatigue life through experimental and modeling approaches. Axial low-cycle fatigue tests on base metals and RFW specimens at varying friction pressures show a direct correlation between friction pressure and fatigue strength coefficient. The fatigue life was modeled using the Coffin-Manson equation and refined with an Adaptive Neuro-Fuzzy Inference System (ANFIS) for enhanced prediction accuracy. Results demonstrated that higher friction pressure improves fatigue life and weld strength. This research provides insights into optimizing RFW parameters for better fatigue performance of carbon steel joints. The integration of empirical, experimental and ANFIS-based modeling offers practical guidelines for selecting welding conditions, improving durability and reducing fatigue testing costs in industrial applications.