The highlights of this release are "Create SOP from file" and "Specify reference SOP in video analysis (beta)". We've built a workflow so that many customers can leverage their existing SOP assets currently managed in Excel into Dive's video SOP. At the same time, we're adding QR codes to output files, improving layouts for mobile and tablets, and refining viewing records and usage reports.
① Create SOP from file (Excel/Word/PowerPoint support)
We've expanded "Create SOP from file," which previously supported PDF only, to also support Excel (.xlsx/.xlsm/.xls), Word (.docx/.doc), and PowerPoint (.pptx/.ppt). You can upload existing Excel manuals, work standards, checklists, and more, and have AI turn them into SOPs.
- Most customers currently manage SOPs in Excel. The goal of this release is to move your existing assets directly into Dive without re-shooting videos or rewriting content.
- When you upload, the system converts to PDF, and AI generates title, summary, and steps (including questions and pass/fail judgment) all at once.
- In AI processing options, you can fine-tune how content is imported (body text mode: formatted/summarized or transcribed/original text), step split method (per content or per page), tone (command, polite, suggestive, or follow policy), and conversion mode (follow print settings, fit width to 1 page, or 1 sheet = 1 page).
- After importing, editing, sharing, and output flow seamlessly in the same editor as video SOPs.
For details, see Create SOP from file (Excel/Word/PowerPoint support).
② Specify reference SOP in video analysis (beta)
Building on ①, you can now specify existing SOPs or files (PDF/Excel/Word/images) to reference when analyzing video. This addresses the need to refine video manuals based on existing Excel SOPs or to reuse terminology and structure you've already established.
- [Split video to match this document's sections (strict alignment)] checkbox ON: Steps are divided to match the step count and order of the reference file. Title and body text are taken directly from the reference document, with information read from the video added to the body (mode for creating video manuals based on existing work procedures).
- Checkbox OFF (default): Reference file is used only as supplementary material for terminology and context, with step division automatically determined from video.
- Beta in progress: Division accuracy and text incorporation methods are still being refined. If you notice examples that work well or issues, please send feedback to customer-success@episotech.com.
For details, see Specify reference SOP in video analysis (beta).
③ Support for "Take photo and save" when answering questions
You can now take photos as evidence when answering questions. Use this when you need to capture proof on-site during inspections, rounds, or checklist operations. Photos are linked to viewing records and can be reviewed or converted to reports later.
④ Add QR Code to output files (PDF/PowerPoint)
When exporting SOPs as PDF or PowerPoint, you can now embed QR codes for each step. Even after distributing procedures on paper or slides, users can scan the QR to jump directly to Dive's playback screen (with video). Use this as a bridge for "dual operation of print and digital."
Other improvements and bug fixes
- Major layout improvements for mobile and tablet: We've completely reviewed button placement and scroll behavior on narrow-width devices across key screens (SOP playback, editor, AI processing options, etc.). We've also standardized the approach for portrait-oriented tablets as "large smartphones" to make behavior more predictable.
- SOP playback: Player font size now adjustable: Readers can now adjust the font size of text on the playback screen. On-site staff using smartphones or small devices can read SOPs in a comfortable size.
- Text retained even after cut editing: When you cut a video, text associated with that section (SOP body, title, captions, etc.) is no longer lost and carries over to other steps. The experience of "cuts cause text to disappear and require rewriting" is eliminated.
- Refined viewing records and usage reports: Improved display of viewing record detail modals and list tables (filtering and sorting, photo answer previews, card view for mobile). Cleaned up individual usage report presentation as well.
- Simplified "remaining seats" display in User management: Streamlined the remaining-seat summary for permission frames (owner-to-viewer permission frames and guest permission frames) in the User management screen to show available seats at a glance. For more on permission frames, see User permissions and permission frames.
- Fixed AI processing count bug: Corrected cases where AI processing count was incorrectly double-counted.
- Optimized Terminology Glossary page performance: Improved page display and editing responsiveness for teams with large numbers of dictionary entries.
- Fixed issue where video description input disappears quickly: Corrected a bug in the AI processing options "Video description" field where entered text was unintentionally truncated.
- Organized caption and read-aloud settings modal: Aligned the layout of captions, read-aloud, and question mode settings screens for both "Create from file" and "Create from video."