Frequently asked questions relating to the 'Import unstructured documents' functionality.


TABLE OF CONTENTS


General questions about 'Import unstructured documents'

Which content sets have the 'Import unstructured documents' feature?

February 2026:  'Canada - NMS and CMS' and 'USA - USMS'.


What file formats can be imported via the 'Import unstructured documents' feature?

  • Supported:
    • .docx, .pdf, .xml.
    • .zip containing multiples of the file formats above.
  • Unsupported:
    • doc, .rtf, .txt.

Can you import multiple files at once, or do you need to feed them in one at a time?

Yes, separate files should be imported as a .zip file.


How long does the 'Import unstructured documents' process take?

10-15 minutes per file, depending on the complexity of the specification content, depending on:

  • Length of the file:
    • Single file with a small section vs a single file with multiple sections.
    • A ZIP containing lots of single files with fewer than 20 pages doesn’t take as long as files with lots of pages or complexity per page.
  • Complex formatting:
    • Tables.
    • Text formatting (bold/ italics/ underline and so on).
    • Complex nesting of paragraphs.
    • Empty paragraphs (rows).


What should I do to check that my files have been fully imported?

We recommend reviewing the import report attached to the 'Import successful' email. The report contains a 'Similarity (%)' column containing a score. The 'Similarity (%)' compares the plain text between the source document and the imported spec on a paragraph-by-paragraph basis.


What should I do if there is a problem importing my specification?

We recommend reviewing the import report attached to the 'There was a problem importing your specification' email. The 'Notes' column will show the reason why the file failed to import. 

Related article: Problems importing your specification


What happens if the source files contain tables, hidden guidance and images?

  • Tables are imported.
  • Potential exceptions to this include:
    • Complex tables.
    • Tables positioned outside of articles.
  • Hidden guidance is not imported.
    • February 2026: This is in development and expected to be released in March 2026. Hidden guidance will import into 'Organization notes'.
  • Images are not imported or reported on.


Is my data used to train AI models?

No. Your uploaded documents are only used to perform the requested transformation and are not used to train AI models.


When the import process experiences issues


The status shows “Import pending” for a long time. Is that normal?

Yes. Large or complex files can take longer to analyze. If the status shows 'Import pending', the system is still working.


I imported my document 2h ago, but it's still pending.

The import and report generation may take a few hours.


I imported a zip file, but some sections are missing from my specification. Why is this happening?

Verify if you've received an email titled 'There was a problem importing your specification' regarding the missing sections. If not, the import process might still be ongoing. Please be patient a little longer.


There was a problem with my import. The 'Notes' indicate there was a 'A request timeout error'.

Please retrigger your import. If that doesn't resolve the issue, try doing it overnight or early in the morning. If it still doesn't help, consider simplifying it by removing tables or deeply nested paragraphs.


A specification was imported wrongly, with elements mismatched.

Please try importing it again, as results may vary depending on the AI processing involved.


I uploaded a zip file and got several emails.

This is the expected behaviour. Each file in the zip archive is accompanied by a separate email report.


Why did I receive a report even though everything looks correct?

Reports are always sent after processing finishes, even if no issues were found. This way, you get confirmation that the import completed successfully.


What should I do if the report mentions warnings but the import succeeded?

Warnings usually mean minor parsing differences. Check the affected sections in the specification platform and adjust them if needed.

 

Can I cancel an import that is already running?

No, you can’t cancel the import once it has started.

 

Why does processing time vary between similar documents?

Processing time depends on the document’s structure, tables, nested content, and overall complexity, not just the file size.