Skip to main content

Upload a document

DocuDesk treats every file as a typed Document object stored in the OpenRegister Documents register. Uploading a file kicks off the metadata-enrichment pipeline (language detection, keyword extraction, topic classification) and makes the file available to every other DocuDesk flow.

Goal​

By the end you will have uploaded one document, watched the metadata enrichment fill in, and inspected the resulting Document object.

Prerequisites​

  • You have completed Open DocuDesk for the first time.
  • You have a sample document on disk — PDF, DOCX or plain text work; image-only files are accepted but won't trigger language detection without OCR.

Steps​

  1. From the navigation, click Documents and then Upload.

    Documents list with Upload button

  2. Pick a file from your local disk. The upload dialog shows file name, size and detected MIME type before you confirm.

    Upload dialog

  3. Submit the upload. The document appears in the list with a Processing… badge while the metadata-enrichment pipeline runs in the background.

    Document being processed

  4. After processing (usually a few seconds for short documents), the badge clears. Click the row to open the detail view. The right-hand panel shows the enriched metadata — detected language, top keywords, topic classification.

    Document detail with enriched metadata

Verification​

You are done when: the document appears in the Documents list, the Processing… badge has cleared, and the detail view shows at least Language and Topic fields populated.

Common issues​

SymptomFix
Upload fails with "MIME type not allowed"The instance admin restricts allowed types in Settings → DocuDesk → Allowed MIME types — ask them to extend the list.
Processing… badge never clearsThe metadata-enrichment service (Tika / Solr / language-detection) is unreachable — check Settings → DocuDesk → Service status.
Detail view shows empty Language / Topic fieldsThe document is too short for reliable detection (under ~200 words) or contains only scanned images — try OCR before upload.

Reference​