Classic database application.
You say scan paperwork which suggest no OCR and thus a need to store images of said documents.
Database would have 1 table (entity) with 2000 records, each record would have 2 fields (attributes). One field would hold the record identification (e.g. 1234 - BUT make sure you set the datatype as text NOT integer), the other field would hold the image.
MS Access can handle all this and has a fairly friendly GUI to assist in the setup.
Using a Database means you get all the tools built in for backup, repair, replication, sorting, searching (only on record identification field if the docs are images, suggest a third field which contains text key words relating to doc contents - this would be searchable).
Cron
PM if you need more help.