Agriculture productivity maps based on satellite images and machine learning algorithms have become powerful tools for understanding and optimizing agricultural practices. By combining the capabilities of satellite imagery and machine learning algorithms, these maps provide valuable insights into crop health, yield potential, and resource management. Satellite images capture detailed information about vegetation, soil moisture, and other environmental factors that affect crop growth. By training a machine learning model on a large dataset that includes historical yield we generate accurate predictions of crop productivity. This dataset contains information about crop productivity (maize, soybean, sunflower and sugarbeet) in the Vojvodina region (Serbia) for 2019. season.