When a file arrives, Fryri reads it, works out what it is, and files it. You do not sort anything into folders; the filing is Fryri's job.
Reading#
Text is extracted from documents, including scanned PDFs, which go through OCR. Photos and videos are described, so a photo of a receipt or a whiteboard becomes searchable by what is in it. Voice recordings are transcribed.
Filing#
Each document gets a type (invoice, contract, ticket, and so on), a date, and the names, amounts and details found inside it. That structure is what makes questions like "how much was the March invoice" answerable.
Where it stays#
The original file stays in your library. You can open it, download it, or delete it at any time. Deleting is covered in delete things, or everything.
If reading a file fails, Fryri retries on its own and the file shows its status in the library until it succeeds.