The rapid expansion of Earth observation data from satellites, aerial platforms, and cloud-based archives has created unprecedented opportunities for large-scale environmental monitoring and analysis. At the same time, the volume, complexity, and velocity of these data demand new approaches that are both scalable and accessible. This presentation focuses on open-source GeoAI tools designed specifically for Earth observation applications, highlighting how artificial intelligence can be integrated with geospatial data science to enable efficient, reproducible, and transparent workflows. The talk introduces the GeoAI Python package and its ecosystem, which support a wide range of Earth observation tasks, including land cover classification, object detection, semantic segmentation, and change detection. Through real-world examples, participants will see how open-source tools can streamline data access, model development, and interactive visualization. The integration with QGIS and web-based mapping frameworks further enables seamless analysis across desktop, cloud, and browser environments.