For batteries to function effectively, it is crucial to precisely assess their operational internal state conditions, during the course of the charging and discharging cycles. A precise state-of-charge (SOC) estimation can increase battery safety, dependability, and lifespan. This data is fundamental for maintaining the safety and reliability of the battery management systems (BMS). Since a direct measurement is not feasible, other alternative procedures must be used to estimate and infer battery state information. To distinguish this work from existing knowledge, in this study, a unique method for estimating the SOC of sodiumion batteries (SIBs) is proposed. This fuses a model-based and data-driven approach to develop a hybrid framework. The datadriven model complements the mechanistic battery model which offers domain knowledge and this serves as the foundation for the proposed hybrid strategy, thus improving the interpretability and robustness of battery modeling. The performance of the proposed hybrid method is evaluated through empirical testing using a range of validation techniques and evaluation metrics, confirming its reliability across different assessment frameworks. The Mean Absolute Error (MAE) showed that on average, the hybrid model is off by about 2.94 % and an R2 of approximately 0.87 showed that about 87 % of the variation in the true SOC is explained by the model.