Uploading documents

PDF and Word are read in the browser before the content is sent for analysis. If the PDF is scanned with no text layer, text recognition runs instead.

Long documents are split into parts and analysed in several rounds. The findings are then assembled, and contradictions between different parts of the document are stated explicitly.

You always see how much of the document was actually covered, how many parts were used, and any warning that something was left out.

Your files stay in your workspace until you delete them, and they are not used for anything other than answering your own analyses.