Employee turnover constitutes a substantial challenge for organizations, exerting an impact on operational continuity and overarching productivity, thereby incurring concealed costs. Consequently, estimating the likelihood of employee turnover emerges as a valuable endeavor for Human Resource Departments. Such forecasting empowers these departments to proactively undertake measures aimed at averting such occurrences, particularly when staff is deemed pivotal to organizational success. Our research endeavors to explore the predicting capacity inherent in personality assessment test scores for anticipating turnover tendencies among candidates and employees. Utilizing a data-driven methodology, our study aspires to forge a robust correlation between personality traits and turnover. Furthermore, our modeling of turnover probability incorporates the influence of demographic information (e.g., gender, age, and education level), as well as performance and work-period metrics. Utilizing several machine learning methods, we cast the problem of predicting turnover tendency of an individual as a binary classification problem and train separate classifiers for candidates and current employees. Our experimental findings indicate that the differentiation between managers and subordinate employees proves advantageous in predicting turnover tendencies, given the requirements of distinct psychological traits at various hierarchical levels.