More than 15 years of experience on Machine learning approches to process satellite imagery. 1) Development of the first system for monitoring deforestation in South America in near real-time, called “Terra- i: an eye on habitat change” (www.terra-i.org)  in collaboration with the Centre for Tropical Agriculture (CIAT), King’s College London and Nature Conservancy. 2) Development of a system for tracking deforestation at pan tropical level. 3) Analysis of current and historical satellite images to develop tools for assessing zero-deforestation commitments in Vietnam using Deep Learning models. 4) Development of super-resolution maps of land use. 5) Development of few-shot learning methods for object and time-series pattern matching in aerial drone images.  6) Development of the winner solution of the ACM GISCUP 2023  challenge regarding the detection of lakes formed from melting ice in Greenland.