Modern stress management techniques have been shown to be efective, particularly when applied systematically and with the supervision of an instructor. However, online workers usually lack sufcient support from therapists and learning resources to selfmanage their stress. To better assist these users, we implemented a browser-based application, Home Sweet Office (HSO), to administer a set ofstress micro-interventions which mimic existing therapeutic techniques, including somatic, positive psychology, meta cognitive, and cognitive behavioral categories. In a four-week feld study, we compared random and machine-recommended interventions to interventions that were self-proposed by participants in order to investigate effective content and recommendation methods. Our primary fndings suggest that both machine-recommended and self-proposed interventions had signifcantly higher momentary efficacy than random selection, whereas machine-recommended interventions offer more activity diversity compared to self-proposed interventions. We conclude with refections on these results, discuss features and mechanisms which might improve efficacy, and suggest areas for future work.