Content and media problems

​

Cloud

• Confirm the source uses a supported format and is not password protected.
• Convert older office files to a current format before uploading.
• Use a text-based or higher-quality scan when extraction misses content.
• Check that important data is clear and consistently labeled in the source.

​

Self-hosted

• Confirm the upload and persistent-data directories are writable.
• Check document-processing logs for parser, OCR, file-size, or timeout errors.
• Verify image and search provider credentials independently from the text provider.
• For host services such as Ollama, use a hostname reachable from the container rather than container localhost.
• Convert .doc, .ppt, and .xls files to modern equivalents when parsing fails.