This paper proposed TPOT-RBFN, a novel hybrid approach that integrates Tree-based Pipeline Optimization Tool (TPOT) with Radial Basis Function Networks (RBFN) to address the challenges of automated predictive analytics. The methodology combines the automated pipeline optimization capabilities of TPOT, which employs genetic programming to select optimal preprocessing steps and model configurations, with the robust function approximation properties of RBFN, a neural network architecture known for its ability to model complex data patterns. The RBFN component is characterized by its hierarchical structure, where radial basis functions in the hidden layer enable non-linear transformations, while the output layer linearly combines these transformations to produce predictions. The TPOT framework optimizes critical RBFN hyperparameters, including the number of basis functions, their centers, and widths, as well as the weights connecting the hidden and output layers, thereby enhancing the model’s predictive accuracy. This integration not only automates the traditionally manual and error-prone process of model selection and hyperparameter tuning but also leverages the complementary strengths of genetic programming and neural networks to handle high-dimensional and non-linear datasets effectively. Experimental results demonstrate that TPOT-RBFN outperforms conventional methods in terms of both accuracy and computational efficiency, making it a scalable solution for real-world predictive analytics tasks. The proposed method is particularly significant for domains requiring interpretable yet powerful models, as it balances automation with performance while reducing the reliance on domain expertise. Furthermore, the flexibility of TPOT-RBFN allows it to adapt to diverse datasets, offering a generalizable framework for advancing predictive modeling in both academic and industrial applications.