Start building the Plex database from a file (csv, or XML, ...)

Plex by default doesn’t care about your pre-downloaded metadata (most of them).
It works better if you let it do that itself.
Of course to download the right metadata, it must recognise which movie it is. Which can be trickier than most ppl think, because there are sometimes dozens of movies with a similar (or even identical) title.

So Plex depends on you helping it by providing at least a correct file name which includes the release year.
There is one thing your media manager can help with: let it write an XBMC/Kodi compatible .nfo file into the subfolder of every movie (you do use a subfolder per movie, do you?).
This little .nfo file is a XML document which should contain the IMDB ID of your movie. If this is present, it helps Plex a lot to identify and successfully ‘match’ the movie.

There is another thing, your media manager should be able to do to help: change the file and folder name of the movie to Plex’s requirements.
Alternatively, you can use other software like e.g. Filebot

You can also use the .nfo files directly for gathering metadata. You can install the .nfo importer agents which read all the other infos from the .nfo file and put them into Plex’s database.