Electric vehicles (EVs) are progressively recognized as sustainable replacements to internal combustion engine vehicles. Estimating the state of charge (SoC) accurately in lithium-ion battery systems is essential for safe and efficient electric vehicles. This paper provides a concise overview of machine learning techniques employed for SoC estimation and thermal management in EV batteries. The discussed machine learning methodologies encompass Artificial Neural Network (ANN), Linear Regression (LR), Support Vector Machine (SVM), Gaussian Process Regression (GPR), Ensemble Boosting (EBo) and Ensemble Bagging (EBa) in electric vehicles. Additionally, the paper delves into innovative approaches for Battery Thermal Management Systems (BTMS), specifically focusing on liquid-cooled batteries integrated with heat pump air conditioning systems (HPACS). The Particle Swarm Optimization (PSO) algorithm is frequently utilized to optimize hyperparameters for the Support Vector Regression (SVR) model in the context of the BTMS. Moreover, the review presents a SoC estimation technique that employs a neuro-fuzzy system based on subtractive clustering. A unique U-shaped liquid cooling method aimed at enhancing thermal safety and reducing weight in prismatic battery cells is also explored. Notably, the GPR model demonstrates high precision in forecasting battery temperature. Overall, the review underscores the importance of precise SoC assessment for enhancing energy efficiency, extending driving range, and prolonging the lifespan of EV bateries.