It takes emails – PST files filters them (by date), keyword searches them, and de-dupes them. It then outputs PST files in several different formats, e.g you can produce a single PST file for all the users/custodians you loaded in, or you can produce separate PST files for each custodian. It can also handle horizontal/global and vertical/user based de-duping procedures.
Trident can also output basic load files, though don’t try using it to load data into Introspect or the like.
Wave will also handle standard e-files (DOCs, XLS), and do the de-duping and all the good stuff listed above. It can also handle NSF files (outputting NSF files), but unfortunately it can not keyword search attachments for NSF files – they need to be converted to PST files if that is required. But Trident admit this (though its not widely advertised) so its not a flaw, its a documented feature!
In short, its a great tool.
The primary flaw with it are that it can only handle PST files, i.e it can’t handle DBX, OST, or encrypted PST – they need to be converted/decrypted if required. Tools like FTK can handle all those email formats – but can’t do the de-duping that Trident can.