Indexing, availability, and plan limits

Prerequisites

• A Paperplane workspace on any plan, free or paid.
• Documents uploaded directly, or imported through a connected email account or drive.

Steps

1. Upload a document directly, or bring documents in through an integration: Gmail, Outlook and IMAP mailboxes sync on a recurring schedule, while Google Drive, SharePoint, Dropbox and file servers (FTP, FTPS or SFTP) are imported once per folder you choose.
2. Allow a short time after upload or import for background processing — Paperplane extracts the text (including OCR for scanned documents, in German and English), splits it into page-aware passages, and indexes them before the document becomes searchable through AI archive search.
3. Ask your question once processing has finished. Directly uploaded documents and integration-imported documents are searched together in the same archive — there’s no separate index per source.
4. For questions with a specific, findable answer — a clause, a figure, a date, a condition — expect a grounded answer with source references once the relevant document is indexed.
5. For questions that ask for totals or counts across your whole archive, such as “how many invoices do we have in total,” use quick search or your own records instead. Archive search retrieves and reasons over a bounded set of relevant passages; it does not deterministically scan every document in your archive to produce a count.
6. Track your monthly usage: the AI archive search window shows how many questions you have left this month. Each user has their own monthly allowance, which depends on the workspace’s plan — see pricing for current allowances.
7. If you need more questions, move the workspace from the Free plan to a paid plan, which includes a larger allowance per user, or wait for the reset at the start of the next calendar month.

Expected result

Newly uploaded or imported documents become searchable through AI archive search a short time after processing finishes. Every plan carries a monthly AI-question allowance per user. Specific, content-level questions get grounded answers; exhaustive counts and totals across the whole archive are outside what archive search is designed to do.

Common problems

• A document I just added isn’t showing up in AI answers. It’s likely still finishing background processing. Wait a short time and ask again.
• A document imported through an integration doesn’t appear in AI answers. Check that the import or the last sync actually completed and that the document’s folder and file type were included.
• I asked for a count or total and got an unsatisfying answer. This is expected — archive search works from a set of retrieved passages, not a deterministic scan of every document. Use quick search or your own records for totals across the archive.
• I’ve used my monthly AI archive questions. Check the allowance per user on the pricing page; you can move to a paid plan or wait for the reset at the start of the next calendar month.