How do you handle database scaling when transitioning from a localized data model to a globally distributed cloud architecture for Verkada?
