This study presents an advanced petrophysical evaluation of the Sarvak Formation in one of the major supergiant oil fields in Southwest Iran, achieved by integrating Nuclear Magnetic Resonance (NMR) log data with conventional well logs. NMR measurements from Well-A were analysed to extract critical reservoir properties—including total and effective porosity, and volumes of bound and free water—which significantly enhanced the accuracy of the petrophysical model. A multi-resolution graph-based clustering (MRGC) algorithm was developed to estimate NMR-derived parameters from conventional logs for the adjacent Well-B, where NMR data were unavailable. The MRGC model utilised gamma-ray, acoustic, density, neutron, and photoelectric logs to predict total and effective porosity, clay-bound water, irreducible water saturation, and other NMR-related parameters. The model was calibrated using data from Well-A and subsequently applied to Well-B, enabling NMR-informed petrophysical characterisation in the absence of direct measurements. The optimised petrophysical model demonstrated consistent reservoir characteristics across both wells. Average total porosity was 10.7% in Well-A and 12.2% in Well-B; effective porosity averaged 10.2% and 11.8%, respectively; clay volume was approximately 3.2% in Well-A and 3.6% in Well-B; and water saturation was 85% and 84%, respectively. Based on cutoff thresholds of 5% porosity, 15% clay volume, and 50% water saturation, net pay intervals were delineated, yielding approximately 31 m of productive zone out of 411 m in Well-A, and 30 m out of 380 m in Well-B. The NMR-augmented analysis provided more precise differentiation of hydrocarbon-bearing zones and proved more cost-effective than traditional log-based methods. This refined petrophysical workflow significantly improves reservoir characterisation, enhances the accuracy of hydrocarbon volume estimation, and supports more informed field development planning.