Summary

• Data science has never been as much about machine learning as it has about cleaning, shaping, and moving data from place to place.

• Here are the important concepts in data management:

  • Sources - how to get training data

  • Labeling - how to label proprietary data at scale

  • Storage - how to store data and metadata appropriately

  • Versioning - how to update data through user activity or additional labeling

  • Processing - how to aggregate and convert raw data and metadata