We can no longer ignore the growing evidence about the negative impact of our human activities on the environment, as we are currently witnessing unprecedented levels of biodiversity decline, climate change and environmental pollution. Extensive environmental monitoring is needed to better understand the spatiotemporal patterns in environmental decline, which can then be linked to behavioral patterns (e.g. land use change, pesticide usage, consumption patterns and groundwater depletion). The combination of earth observation data and deep learning is currently causing a revolution in environmental monitoring, a sector that until recently was mainly supported by in-situ monitoring and expert opinion. The yearly number of research papers that apply deep learning in an environmental monitoring context has risen from less than 100 in 2015 to over 2000 in 2020 (Google Scholar search performed on 28/05/2021 with search string “[deep learning]&[environmental monitoring]”). Over the last decade, the earth observation sector has entered its era of big data, as a wealth of novel satellite data streams have become freely available through the European Union’s Copernicus programme. Concurrent developments in the artificial intelligence sector – specifically the development of deep learning algorithms – have uplifted the operationalization potential of these huge datasets. Within the GEO.INFORMED project, a 4-year project funded by the Belgian science policy, we aim at developing deep learning workflows that can transform Copernicus Sentinel data into the operational information that is needed by environmental policy agencies.