When machine learning is applied in healthcare, exploring the output model is a critical part of research but can also help the healthcare practitioners. We present Health Wellness Profile UI (shortened as HWProfile UI, available at https://hwprofile.eu-de.mybluemix.net), an exploratory interface developed as part of ProACT (https://proact2020.eu), an H2020 European funded project which aims at advancing home-based integrated care through IoT, analytics and a behaviour change framework. This interface has been conceived with the aim of easing the exploration of a holistic model for the person or cohort of patients, based on a Bayesian Network that represents older patients with multimorbidity, with variables scattered on several dimensions (demographics, clinical, behavioural). HWProfile UI demonstrates a novel and interactive way to navigate probabilistic conditional dependencies through a dynamic view, with the goal of being easily understandable by stakeholders coming from different domains (e.g. IT specialists, nurses, psychologists, etc.).