This research paper focuses on the use of an enhanced hybrid time-series forecasting technique for gold rate prediction. The proposed technique is a combination of two models, namely the Auto-Regressive Integrated Moving Average (ARIMA) and Artificial Neural Networks (ANN). The ARIMA model is used to capture the linear components in the data, while the ANN model is used to capture the non-linear components. The proposed model is applied to daily gold rate data from January 2010 to December 2022, and the results are compared with those obtained from the individual ARIMA and ANN models. The experimental results show that the proposed hybrid model outperforms both the individual ARIMA and ANN models, with a Mean Absolute Percentage Error (MAPE) of 0.89%, which indicates a high degree of accuracy in gold rate prediction