The WorkingAge project aims at improving the psycho-physical condition of workers, with a special focus on ageing subjects. In this context, a Decision Support System, based on a hybrid data-driven / model-driven approach, fed with data coming from environmental and wearable sensors, aims to provide personalised advises to the worker. In this paper we briefly present the WorkingAge project and architecture, and then focus on the decision-making pipeline that, starting from raw data, generates the advises.