If you needed to scale your solution to handle 10x the current data volume from the Earth Lab pipeline, what bottlenecks would you hit first?
