QUESTIONS
Does the data lake contain an exact copy of data from source systems? A subset of common dimensions? Why is there risk of deleting source data? Rather than storing an exact copy, why not store in a format for analytics?
Are there guards as to what gets stored in data lake? How about the format it gets stored in?

ANSWERS
Informatica
Visualisation
Common API

|total time to build model - time to understand data| makes sense, because collaboration to make it easier for others to understand data are helpful overall but excluded from metric