Recently, research on solid electrolytes with high energy density and stability has been actively conducted. This study proposes a new exploration method for composite electrolytes formed by synthesizing NASICON, one of the inorganic electrolytes, and PVDF-HFP, one of the polymer electrolytes. For this purpose, 24,181 NASICON materials were created through element substitution and screened based on synthesizability, thermodynamic stability, insulation, mechanical properties, and ionic conductivity. In the process of creating the structure of NASICON and predicting its properties, Graph Neural Network-based Machine Learning Interatomic Potential and machine learning models were used. Finally, the electrochemical performance of the composite produced through synthesis of the screened NASICON material and PVDF-HFP is tested, and the feasibility of the screening process and a new composite electrolyte material are proposed.