Overview
In the PDF SOP creation feature, the "Split method" setting was not being applied correctly during import. Even when "Per content" was selected, the system always processed as "Per page" (1 page = 1 step). Materials with table formats or multiple steps (rows) on a single page (such as standard work sheets) were not being split by row.
Date/time and Status
Fix completed on July 15, 2026.
This issue occurred specifically when uploading a new PDF and converting it to an SOP using AI. SOP creation from Excel / Word / PowerPoint, and re-running from the Materials List ("Run AI processing") were not affected.
Root Cause
When uploading a new PDF, the selected split method (Per content / Per page) was not being passed to the AI analysis process. The system always processed using the default "Per page" setting. As a result, even materials with multiple steps on a single page were being created as 1 page = 1 step.
Fix Details
We corrected the process so that the selected split method is now properly passed to AI analysis when uploading a new PDF. Additionally, we unified the default split method to "Per content" when not specified.
Impact Range
PDFs uploaded and converted to SOPs before July 15, 2026 may have been created as 1 page = 1 step regardless of the split method selection. If you are using table format or multi-step materials, the split may not have been done by row as intended.
Action Requested from Users
- If you want to re-split an already uploaded PDF by row: From the Materials List, select [Run AI processing (re-run)] for the target file and specify [Per content] to re-analyze. Table rows and bullet points within a single page will be split into separate steps.
- After this fix is released, the [Per content] setting will be correctly applied from the moment of PDF upload.
- If the number of affected files is large, we can assist with reprocessing. Please feel free to Contact us.
We sincerely apologize for the inconvenience this has caused. We look forward to continuing to serve you with Dive.