Space weather events related to solar activity can affect both ground and space-based infrastructures, potentially resulting in failures or service disruptions across the globe and causing damage to equipment and systems. Global Navigation Satellite Systems (GNSS) represent one of such infrastructures that can suffer from electromagnetic phenomena in the atmosphere, in particular due to the interaction of the RF signals with the ionosphere. The Ionosphere Prediction Service (IPS) is a project funded by European Commission to provide a prototype platform for a monitoring and prediction service of potential ionosphere-related disturbances affecting GNSS user communities. It is designed to help these communities cope with the effects of the ionospheric activity and mitigate the impacts of these effects on the specific GNSS-based application/service. The IPS development has been conceived of two concurrent activities: the design and implementation of the prototype service and the research activity, which represents the scientific backbone of IPS and is at the base of all the models and algorithms used for the computation of the products. The products are the basic IPS output that translate the nowcasting or forecasting information from the whole IPS system down to the final user. They are fine-tuned to match the different needs of the communities (scientific, aviation, high accuracy, etc.) which the service is targeted to and to warn the GNSS users about possible performance degradations in the presence of anomalous solar and atmospheric phenomena. To achieve this overarching aim, four different blocks of products dealing with solar activity, ionospheric activity, GNSS receiver and system performance figures have been developed and integrated into a unique service chain. The service is available to a set of invited users since July 2018 through a web portal and its provision with all the necessary operations will last 6 months. The prototype will be also ported to the Joint Research Centre (JRC). This phase will be useful to further test the platform, and to assess whether and how a dedicated prediction service for