What to do with the saved files in the Ninox_DB?

The good thing about Ninox is that you can use the apps to directly create photos for the record. As long as you use the app only and also save the database on your own computer or in the iCloud, you don't really have to worry about the database size. As you know, images/videos and this also includes the print PDFs created and saved in Ninox, inflate the database size.
If you now use the Ninox Public Cloud, you can use the new consumption overview (cogwheel in the team menu) to see what volume has accumulated over time. (I should not have done this)
How can you counteract this.
One possibility is to monetarily increase your Ninox capacity.
My approach is to slim down the database by offloading the files. There are a few solutions on the market: NextCloud, Seafile, Baserow etc., which you can also host yourself. I decided on Baserow and use n8n (although it would certainly also work with Make) to outsource the files and only keep a share link to the file in Ninox. A third-party provider like n8n because with Baserow you need a certification key before uploading and you can't upload images directly from Ninox anywhere. With images, you still have the advantage of getting a small preview via thumbnails. In my case, the Ninox DB's where I already use it are considerably reduced in size and the workflow within Ninox also runs smoothly. The system also works with the apps (I use the iPhone to photograph till receipts and automatically enter the amount/date paid into Ninox via an OCR) After the OCR, I store the photo directly and thus keep the DB small.
You can find an example of the functionality for outsourcing and re-importing files in the NX-API Team in the Carbone example DB in the table Images. You can test it directly in the team, because no Ninox API key is needed. However, a Ninox ABO is required, as sharlinks are generated for transfer and this only works with the Ninox cloud versions.

I can also upload a pure NinoxDB to the team for testing, which only shows the upload/import without bells and whistles and includes the n8n Flow.

This is what my n8n Flow looks like for transferring via webhook:
  

This is what the Baserow table looks like:

Since I only use a small number of Ninox databases in production, I have started to store the 'test databases' outside of Ninox in order to load them again when needed. This can be done locally on the hard disk or with Baserow. I recently loaded a 190MB database into Baserow.