Probabilistic Slope Stability Analysis (PSSA) is crucial for understanding the underlying spatial variation in soil parameters which affects slope analysis and design. In recent years, intelligent data-driven systems have become essential for solving complex real-world problems. Hence, this study presents a comprehensive data analysis and prediction framework employing a Cross-Attention Gated BiLSTM model. Data cleaning identifies and removes duplicate soil test records from the dataset and finds the missing values, outliers to ensure data reliability. Subsequently, Exploratory Data Analysis (EDA) is conducted using descriptive statistics and visualizations to uncover patterns and correlations. After extracting, a deep learning-based prediction model, a Cross-Attention Gated BiLSTM is implemented, which leverages bidirectional temporal dependencies and cross-attention mechanisms for enhancing contextual feature learning and improves predictive performance. The model demonstrates significant potential in capturing complex dependencies in sequential data, making it highly suitable for time-series prediction and classification tasks. Implementing Python software with slope stability dataset, the proposed Cross Attention Gated BiLSTM accommodate the actual generation pattern better than existing methods and produce the Mean Absolute Error (MAE), and Root Mean Square Error (RMSE) , R2 value of 0.06, 0.11 and 97% respectively.