Generated Columns in Databricks is a great feature to automatically create surrogate keys and populate columns using generation expressions.
Generated Columns are defined upon table creation. This can either be generated identity columns returning a unique value. Or it can be a SQL expression (generation expression) to populate a column automatically.
This enables engineers to define transformations (using the columns defined in the table) directly upon creating the table. I am not sure, if this is a good idea.
Certainly identity columns and operations on timestamps like extracting year, month and day are a useful tool. The documentation provides some useful examples.
