HISTORICAL DOCUMENTS PROCESSING APP

What it is:
Digitalized-census-forms-tagging desktop app for a family tree/historical research company.
Built to handle almost a million documents stored on a remote NAS, the application is used by multiple team members to work with the documents.
Supports many image files (jp2, jpg, tiff, etc.) and PDF documents - often big files that need to be shown to the user quickly.
How it's being used:
Employees use the app to manage a record for all of the documents - categorizing them, searching and browsing,
tagging (regular text, years, OCR text, in-document positioned texts).
It is also used in local-only mode for other smaller projects.
Tech:
PostgreSQL for the database.
Mutool for PDF support.
LLMs and Kraken for text OCR.
Avalonia | .NET 10 | C#.
WixSharp for the installer.