This article examines the ability of supervised learning methods such as support vector machines, random forest (RF), and neural networks to characterize individuals exposed to volatile organic compounds (VOCs) using VOC measurements collected from a carpentry workshop in a Palestinian village. The analysis reveals significant differences in VOC measurements among the categories of participated individuals, emphasizing the importance of VOCs as indicators. The predictive models achieved high accuracy in identifying gender, smoking status, and age, with the RF model reaching 100% accuracy in age prediction and 98% accuracy in smoking status prediction. In terms of VOC importance, “Toluene.2.4. diisocyanate” and “Propanal” were the top VOCs for predicting age, while “Toluene.2.4..diisocyanate” and “Bromodichloromethane” were crucial for predicting smoking status. For gender prediction, “Toluene.2.4..diisocyanate” and “Ethylbenzene” emerged as key VOCs. The study contributes to the field of environmental sciences and air pollution research, as well as providing practical recommendations for decision-makers. These recommendations emphasize the need to organize carpentry workshops and similar facilities in a way that reduces pollution and safely protects the health of nearby communities. By applying machine learning techniques, this research provides insights valuable for informed decision-making and for advancing business and environmental management practices in related industries.