When you import a PDF, AI reads the content and automatically assembles it into an SOP. You can reuse existing manuals and work standards as Dive SOPs. AI generates the title, body, tips, cautions, and safety notes for each step from your materials.

What is creating SOPs from PDF

A creation method where AI assembles an SOP based on the PDF you import. The generated content includes:

Title and overview of the entire PDF
Division into steps (by default, rows in tables, bullet points, and multiple procedures within one page are split into separate steps based on content)
Title, body, tips, cautions, and safety notes for each step
Questions (if you select "Create questions" in AI options)
Each step displays the relevant page from the original PDF as is

By default, the body is well-organized text. If you want to keep the original wording, select "Text" in the AI options under "Body writing style". One step will not span multiple pages.

Preparation (supported file formats)

In addition to PDF, you can import Excel, Word, and PowerPoint files as is (no need to export to PDF beforehand). For information on importing Office files, see Create SOPs from file (Excel/Word/PowerPoint supported).

Note that PDFs with selectable text (PDFs containing text information) have higher accuracy than PDFs created by scanning images.

How to create

When creating a new SOP

Select "+" in the sidebar, then "Create from file"
Upload a PDF file, or select an AI processed PDF from "Materials" (if you upload, set AI options and run processing. See "Setting AI options" below)
In "How do you want to import the file?", select your import method from "As-is (1 step)", "1 step per slide", "Select slides", or "File-based" (for differences between each option, see the "How to choose an import method" section in Create SOPs from file (Excel/Word/PowerPoint supported))
Review "Step conversion settings" and press "Confirm with this content"
A draft is created, so review and edit the content, then publish it

When adding steps to an existing draft

You can also import PDFs from the SOP editing screen to add steps. Use this when you want to insert PDF content into an existing SOP. When adding a step, select "File" and choose a PDF, and you can select from the same import methods: "As-is (1 step)", "1 step per slide", or "Select slides". For PDFs that haven't been AI processed, only "As-is (1 step)" can be selected.

Setting AI options

When you upload a PDF, the AI processing options open. Explanations appear when you hover your mouse over the explanation icon next to each item or over option buttons. The settings from your last selection are already filled in.

File description

Telling AI briefly what kind of PDF it is will increase the accuracy of the title and body. Only a description specific to this file is needed. If you have common terminology or industry-specific assumptions shared across your organization, you can register them once in Terminology Glossary and they will be reflected in the analysis without needing to enter them each time.

Step split method

Per content (default): Rows in tables, bullet points, and multiple procedures within one page are split into separate steps. Best for materials with multiple steps on one page
Per page: One page becomes one step. Best for slide-style PDFs with mainly diagrams and photos

Text style

Imperative (~do): The SOP standard
Formal (~do this): For new employee training or SOPs to submit to business partners or auditors
Suggestive (~let's do): For educational and training purposes
How to write: When your team has defined how to write SOPs, match that wording

Advanced settings

Body writing style: "Structured" (default) reads the content and organizes it into work content, tips, cautions, and safety notes. "Text" keeps the original wording as is, faithfully transcribing table rows and bullet points. Best for materials that prioritize the original text, such as regulations, checklists, and evaluation sheets
Recognition Language: Selecting the language of text in the PDF will increase the accuracy of structuring and supplemental generation
Create questions: AI creates questions to present to workers based on the material content. When enabled, you can select "Question type" (On-site check/Comprehension test). When using "Text", this cannot be selected, and only questions already written in the original material will be imported

For detailed explanations of each item, see the "Details of AI processing options" section in Create SOPs from file (Excel/Word/PowerPoint supported).

Review and edit the generated SOP

When processing is complete, a draft is created. Each generated step can be edited like a normal SOP.

Edit title, body, tips, cautions, and safety notes
Delete unnecessary steps or rearrange their order
Add videos, images, or AR to steps

Review the content and edit as needed, then publish it. When editing again, you can access the target SOP from the "Drafts / Approvals List".

Tips for successful creation

Enter a file description: Just adding a note like "Inspection SOP for XX equipment" will improve accuracy
Choose a split method that fits your material: For materials with multiple steps on one page, "Per content" is recommended; for materials organized like slides with one task per page, "Per page" is recommended. If the steps don't split as expected, you can change the split method and re-analyze by selecting "Re-run AI" in the materials details
Use PDFs with text information: PDFs with selectable text have higher accuracy than PDFs created by scanning images

Notes and limitations

This article explains PDF import. Excel, Word, and PowerPoint can be imported from the same screen
When using PDFs in "Materials", you cannot create if AI processing is not complete. In that case, select "Run AI processing" in "Materials" and try again
PDFs with restricted printing, editing, or copying (those that don't require a password to open) can be AI processed as is. PDFs that require a password to open cannot be AI processed, so please prepare a file with the password removed