The collation determines how the database engine compares, sorts, and matches text data of a column.
collation is not yet a managed metadata. Tabulify don't know the column collation.
This has an impact on the sequence generator for character data.
By default, the generated data does not include uppercase letter to have true unique value for case-insensitive collation.
Why?
With a case-insensitive collation (The default of SQL Server, ie SQL_Latin1_General_CP1_CS_AS):
And the generated would not be truly unique in the sequence.