The integration of machine learning continues to garner considerable attention due to its efficiency and reliability. Automated machine learning has revolutionized artificial intelligence by streamlining the development of models capable of addressing predictive challenges. Despite advancements in traditional approaches, there remains a pressing need to enhance model effectiveness and performance. This paper introduces TPOT-RBFN, a novel hybrid framework aimed at improving model performance by leveraging genetic programming (GP) to integrate radial basis function networks (RBFNs). The proposed approach automates the exploration and optimization of RBFN structures and parameters through GP, significantly reducing manual interventions and enhancing predictive capabilities. A real-world dataset was utilized for training, testing, and validating the model in a Python environment. The evaluation metrics included accuracy, precision, sensitivity, specificity, F1-score, and ROC-AUC. Experimental findings on benchmark datasets underscore the superior performance and robustness of TPOT-RBFN compared to standalone RBFNs and other hybrid techniques. In summary, TPOT-RBFN emerges as a powerful tool for predictive analysis, particularly in early medical diagnosis within the healthcare domain, by harnessing genetic programming to refine RBFN-based methodologies.