Fivetran’s SQL Server / Azure SQL Managed Instance destination currently applies a case-sensitive binary collation to replicated string columns instead of honoring the destination database default collation.

In our Azure SQL Managed Instance environment, most existing databases use SQL_Latin1_General_CP1_CI_AS, so this creates constant friction when joining Fivetran-landed tables to application tables, warehouse objects, and cross-database datasets. We are frequently forced to add COLLATE clauses in joins and predicates just to make normal SQL development work.

This causes:

• extra development and maintenance overhead
• recurring collation conflict errors
• less efficient join logic when collation coercion is introduced
• inconsistency with the rest of our SQL Server estate

Requested enhancement:
Please provide a supported option for SQL Server / Azure MI destinations to either:

• honor the destination database default collation, or
• allow destination-level configuration of collation behavior, including a case-insensitive option

This would make Fivetran-landed data much easier to integrate into existing enterprise SQL Server environments and reduce ongoing technical debt.