​This study presents the first high-resolution biostratigraphic and evolutionary analysis of the genus Somalina from the Lapoie and Kelestan sections (Interior Fars, Iran), based on 600 thin sections from the Paleocene-Eocene Jahrum Formation. Analysis of this extensive dataset documented the occurrence of Somalina hottingeri, S. praestefanini, and S. stefanini within the studied sections (SBZ13–SBZ15), contributing to the biostratigraphic characterization of the middle Eocene succession in the Zagros Basin. Geochemical analyses, including X-ray diffraction (XRD), cathodoluminescence (CL), and scanning electron microscopy with energy-dispersive spectroscopy (SEM-EDS), reveal a dominantly calcite-dolomite mineralogy (50–70% calcite, 20–40% dolomite) with minor siliciclastic input, indicative of shallow marine deposition under periodic hypersalinity. Diagenetic overprints, such as pyritization and dolomitization, reflect fluctuating redox conditions during burial. Machine learning algorithms (Random Forest, Support Vector Machines, and Artificial Neural Networks) were employed to evaluate hydrocarbon potential, achieving 95% predictive accuracy by integrating microfacies data and major element geochemistry as key features. The study bridges critical gaps in understanding the paleoenvironmental controls on Somalina distribution, emphasizing the role of tropical Tethyan conditions in shaping evolutionary trends. This multiproxy approach demonstrates the efficacy of integrating traditional biostratigraphy with advanced geochemical and computational techniques to resolve complex depositional histories in carbonate systems. The findings refine the chronostratigraphy of the Jahrum Formation and highlight its reservoir potential through detailed diagenetic and compositional characterization. This work establishes a novel, high-resolution framework for the eastern Tethys and contributes to broader discussions on Eocene carbonate platform dynamics, setting a precedent for data-driven paleoenvironmental reconstructions in sedimentology.