The quality and quantity of Earth Observation (EO) data is increasing at a high rate. However its volume makes handing the data challenging, especially as the bandwidths between data centers do not grow accordingly. In your PhD, you will develop AI methods to enhance the accessibility and analysis of a vast amount of high-resolution, multi-source EO data. You will research ML methods of encode and compress high-volume EO data into practical, much smaller representations and scale them to large datasets on our supercomputers. These techniques are designed to efficiently exchange massive geospatial data for few-shots learning, and facilitating real-time searchers on petabyte-scale data. This research will take place within the scope of the recently funded EU project, Embed2Scale.