Extend Multimodal Image Understanding to Enterprise Search Indexed Knowledge (Standalone Images and Images Embedded in PDF/PPTX/DOCX)

We are deploying Moveworks as an enterprise AI assistant. Many of our existing manuals and knowledge articles contain critical procedures only as screenshots (annotated UI captures, step-by-step images) or as diagrams embedded in PPTX, PDF, and DOCX files.
 

We are glad to see that Brief Me now supports multimodal understanding of images inside uploaded files (Product Update, Idea: Multimodal Support for Brief Me). However, for knowledge that is indexed through Enterprise Search (KB articles and files from connected systems), our support case confirmed that Moveworks does not perform OCR on images. The recommended practice is to add a text description of any important image content to the article body (Writing AI-Ready KB Articles).

This creates a real adoption barrier for organizations with image-heavy knowledge:

• Existing manuals often hold key procedures only in screenshots. Rewriting them at scale is costly and slow.
• Questions whose answers exist only in images, whether standalone or embedded in PPTX/PDF/DOCX, cannot be answered from indexed knowledge.
• Content owners must maintain duplicate text next to every image, increasing maintenance effort and the risk of inconsistency.

Image understanding for enterprise knowledge is also becoming a baseline expectation across enterprise AI assistants and search products, so closing this gap matters for side-by-side evaluations.
 

Requested capability

1. Extend multimodal image understanding to Enterprise Search ingestion, so that images embedded in PDF, PPTX, and DOCX files, and in KB articles from connected systems, are understood and used for answers.
2. Support standalone image files (JPG/PNG, etc.) in indexed knowledge sources.
3. An admin setting to enable or disable image ingestion per knowledge source, for governance and cost control.
4. As an interim step, clear documentation of supported file formats and limitations for Enterprise Search.
    

Questions

• Is extending multimodal understanding from Brief Me to Enterprise Search on the roadmap?
• When will Brief Me’s multimodal understanding be available in the Japan region? (The product update mentions US, CA, and EU first.)
   

Expected benefit

• Faster time-to-value, since existing content can be used as-is
• Better answer coverage for procedural, UI-based questions
• Lower authoring and maintenance burden
• A stronger competitive position in enterprise evaluations