Upload and read a document
From a scanned PDF to text you can actually search.
The path of a document
Upload
Into the document pool, or onto the engagement it belongs to. The accepted format is PDF — scans and photographs go in inside a PDF.
It gets read
The text is extracted, Persian numerals and tables included. This takes time, and how long depends on the page count and the scan quality.
You review it
The output is reviewed by a human before it enters any work. No figure gets in without that step.
It becomes searchable
After a clean run the document is indexed and its content can be used in answers — each one citing its source page.
What the document card shows
Each document has a card carrying its generated summary on the card itself, not only as a hover tooltip. Where there is no summary, the title takes its place.
Status badges and tags live on the same card, and every action — review, download, re-index, delete — is collected behind one menu. Actions that are impossible at that moment are simply absent; while a document is still processing, review and download are not in the menu at all.
One exception: lost pages are reported on the card itself, not inside the menu. State should never hide behind a menu.
If the result is poor
Scan quality is the biggest factor. Skewed scans, low light and stamps over text are the hardest cases.
Handwriting is not supported.
If a document read badly, send us that document — it is the only way to see the problem on a real sample.
Read this next
- DocumentsReview a document page by pageSee the extracted text beside the original image, and correct it.
- DocumentsWhen a page fails to readTotal loss, a few missing pages, and the difference between the two kinds of retry.
- DocumentsFormats and limitsWhat is accepted, what is not, and where the caps are set.
Still stuck?
If the answer wasn't here, write to us. It is a real inbox and it gets answered.