Explainable AI-Based Framework for Predicting Mental Health Vulnerability in Technology Professionals

The technology sector continues to grapple with an alarming rise in psychological health concerns such as stress, anxiety, and burnout. These challenges are often exacerbated by high job expectations, isolation due to remote work, and insufficient mental health infrastructure. This research uses supervised learning approaches—specifically Logistic Regression and Random Forest classifiers—to develop a predictive system for identifying mental health risks among professionals. By analyzing over 1,700 anonymous responses from the OSMI Mental Health in Tech Survey, this study explores patterns tied to occupational and demographic factors. The models are evaluated on metrics like accuracy, recall, precision, F1-score, and ROC-AUC. While Random Forest shows better predictive capabilities, Logistic Regression offers clarity in interpreting key influencing variables. The inclusion of SHAP enhances transparency and supports evidence-based recommendations for improving employee well-being in the tech domain.