Accurately estimating the remaining useful life (RUL) and capacity of lithium-ion batteries essential for optimizing the performance, safety, and efficiency of various applications, including electric vehicles, portable electronics, and renewable energy systems. Traditional methods for predicting battery health often fall short due to the nonlinear and intricate degradation patterns of lithium-ion batteries. This research paper introduces an innovative machine learning-based approach to enhance the prediction accuracy of RUL and capacity for lithium-ion batteries. By leveraging comprehensive datasets from a custom-built battery prognostics testbed at the NASA Ames Prognostics Center of Excellence (PCoE), which includes real-time measurements of charge, discharge, and electrochemical impedance spectroscopy across different temperatures. This study aims to significantly improve battery health monitoring and prognostics