This study presents a multidisciplinary framework for advancing stress detection by integrating Internet of Things (IoT) capabilities with Galvanic Skin Response (GSR) technology. Leveraging IoT for real-time data acquisition and analysis, we enhance GSR sensor functionality. Contributions from clinical psychology focus on elderly populations, providing insights into age-specific stress indicators and mental health correlations. Physics principles optimise sensor accuracy and data fidelity, while computer science provides the framework for data processing, machine learning models, and IoT infrastructure. The linguistic analysis supports psychologists’ recommendations, confirming that GSR readings are best when communication stabilises with subjects. This interdisciplinary approach aims to develop a comprehensive system for effective stress monitoring and management. Our results demonstrate that advanced machine learning models, notably the Random Forest model, achieve high predictive performance in analysing stress levels from GSR data. The confusion matrix and classification report validate its efficiency in accurately distinguishing different stress levels. Feature importance analysis reveals that the GSR Stress Value is the dominant predictor, contributing approximately 90% to stress classification, with age having a minor influence of around ∼ 10%. Gender and state contribute negligibly to stress classification. However, a strong positive relationship between GSR and stress levels is confirmed by a correlation analysis with a coefficient of 0.93. Gender-based differences were minimal, though females exhibited slightly higher stress levels in extreme cases, while younger individuals showed greater fluctuations in stress variability. The findings confirm that GSR data is highly reliable for identifying stress levels and analyzing workplace stressors, offering a datadriven approach to stress analysis and management with significant implications for mental health research and wellbeing strategies.