Upload documents and AI extraction
How the document pool works: uploading files, the automatic analysis that reads each document, duplicate detection, folders, and reminder dates.
Where documents live
Every document belongs to the organisation and the active compliance profile, not to a single product. The Documents page in the sub menu is the pool: folders, a flat view, search, and filters on document kind and compliance area. A document can be linked to any number of products and can fulfil any number of requirements.
Uploading
Open Documents and use the upload action, or drop files on the page. What you can upload is the documentation you already hold: test reports, certificates, Declarations of Conformity, safety data sheets, technical drawings, labels, invoices from test houses and product overviews. Files are stored per organisation and count towards the storage shown under Settings › Plan & Usage.
Before the upload starts, the platform checks for files you already have. Duplicates are listed with the option to skip them (recommended — the existing document keeps working and is not analysed again) or upload anyway. Uploads then run one by one; each file appears in the pool as it finishes, and you can close the window at any point.
You can also upload directly from a product’s Documents tab. Those files go into the same pool and are linked to that product right away.
The analysis
Each new document is queued for analysis. The platform reads the file and works out:
- what kind of document it is (test report, DoC, certificate, invoice, product overview, …),
- which compliance area it belongs to,
- which product identifiers, names or model numbers it mentions,
- which standards it cites, and
- any dates it carries, such as an issue or expiry date.
The status chip on each document shows Analysing, Analysed, Pending review, Analysis failed or Skipped. A failed analysis can be retried from the chip; a document you do not want analysed (a scanned invoice, say) can be skipped. Large batches are analysed in bulk and arrive all at once rather than one by one.
The analysis is what makes matching possible; see Matching documents to products, and share links.
Folders and linking
Folders are optional. Their one special power is inheritance: link a folder to a product and every document in it, and in its subfolders, is linked to that product too. Moving a document into another folder changes what it inherits. Deleting a folder moves its contents to the parent folder; it does not delete documents.
The “Unlinked” view lists documents not connected to any product — useful after a large upload.
Reminder dates and expiry
When you upload a document to a product you can set a reminder date, for example the expiry of a certificate or the date a test must be repeated. The dashboard’s “Upcoming expiry” panel lists documents due within 30 days, and the Notification Center sends “Document reminder” and “Document expired” notifications. Documents without a date are counted as untracked so you can see what still needs one.
Deleting
Deleting a document removes it from every product and requirement it fulfilled and cannot be undone. If a document is merely superseded, upload the new file and unlink the old one from the requirement instead — the history of what was linked stays intact.
If a document has sat in Analysing for more than an hour, send a request with the file name.
Was this helpful?
Thanks — your feedback helps us improve this article.