Tubular structures play a crucial role in renewable energy and the oil and gas industry, particularly in offshore applications. Over time, these structures face significant load, emphasizing the importance of addressing the degradation of tubular joints for sustained operation. Carbon fibre-reinforced polymers (CFRP) offer promising solutions for rehabilitation, yet existing literature primarily focuses on SCF at the crown/saddle points, which is sufficient only for determining fatigue life under uniplanar loads, leaving a gap in SCF along the weld toe. This study aims to bridge this gap by investigating the fatigue design of CFRP-reinforced tubular KT-joints subjected to axial loads at 24 positions along the weld toe. Our research highlights the remarkable potential of CFRP in reducing stress concentration factors (SCFs) in KT-joints, with the degree of reduction correlating with reinforcement layers and elastic modulus. We also uncover the critical role of fibre orientation in optimising stress distribution, particularly when wrapping CFRP around the brace axis with fibres perpendicular to the weld toe. Through 1679 simulations encompassing various geometric and reinforcement configurations, we analyse stress fields at the chord-brace interface under axial compression. Leveraging this data, we employ artificial neural networks to develop empirical models, enabling rapid estimation of CFRP's impact on fatigue life. These models provide precise approximations of hot-spot stress (HSS) in CFRP-reinforced KT-joints under axial load, with less than a 10% difference from simulation results, facilitating accurate fatigue life predictions akin to established methodologies for unreinforced tubular joints.