For one of the requirement, there is a need to integrate to an external database system (i.e. Oracle). The NLS_CHARACTERSET is set to Oracle Unicode character set AL32UTF8
The application has multi-language support (English, French, Dutch). As the application will have multiple data import from CSV and will also export the datasets in CSV file. Can there be any character set conversion loss (UTF-8 and AL32UTF8), which should be taken into consideration?
Although AL32UTF8 Oracle Database character set is based on the Unicode UTF-8 character set, but both of them have different maximum character widths.
Having a look at the default configurations recommended: https://success.outsystems.com/Documentation/11/Setting_Up_OutSystems/Default_Platform_Server_and_database_configurations#Oracle
It says that the charset config should be the one you already have your database on.
Moreover, you have this post from Oracle's community: https://community.oracle.com/thread/3514820?start=0&tstart=0
With all this said, I would assume you won't have losses due to conversion.
Nonetheless for yourself to be confident you should test it to make sure that the languages you plan to support are correctly stored and showed.