A profiling-first workflow for an unfamiliar CSV. It establishes what the file actually contains — types, nulls, cardinality, outliers, candidate keys — before anyone charts anything, because most bad analysis is a chart drawn over an unexamined column.
When to use it
The fi