Abstract
Just-in-time adaptive intervention (JITAI) is a framework used to provide personalized and context-dependent interventions to a user. To fully integrate a JITAI into a user's context, the intervention developer needs to ensure that the user is responding, while still being in the same context. Consequently, they need a context-aware notification management system (CNMS) to accurately time the sending of interventions. This research aims to study smartphone sensor based CNMS for JITAIs in a behavioral change and health context. In this work, we outline the various studies - completed or underway - from a smartphone based digital coach providing interventions and collecting passive sensing data. This data will then be used to train machine learning models, which are finally evaluated in a verification study in the field.