<p>This study presents a machine learning (ML)-driven framework for optimizing the design of polyaniline (PANI)-based supercapacitors, a promising material for energy storage. The framework integrates experimental data and advanced optimization techniques to systematically investigate the effects of key fabrication parameters such as mesh size, deposition cycles, voltage window, and operating potential on performance metrics, including specific capacitance, energy density, and internal resistance (IR) drop. Two ML models, gradient boosting regressor (GBR) and random forest regressor (RFR), were trained on 145 experimental datasets, achieving high predictive accuracy (R² &gt; 0.95). Bayesian optimization (BO) was used to identify the global optimum fabrication conditions, resulting in significant reductions in mean absolute percentage errors 15% for energy density, 20% for capacitance, and 7% for IR drop. Feature importance analysis, including SHAP and partial dependence plots, revealed that mesh size and operating potential were the most influential parameters, aligning with electrochemical theories on ion transport and polymer growth. This integrated approach not only accelerates the design and optimization of PANI-based supercapacitors but also minimizes material waste and provides a scalable, interpretable pathway for ML-driven optimization in the development of high-performance, cost-effective energy storage devices.</p> Graphical abstract <p></p>