New software tools enable practitioners to process imagery from consumer drones into geospatial tree-level forest maps, including species predictions, leveraging computer vision. Applications in California achieve overstory tree detection accuracy over 80% and species classification accuracy over 85%.