Generated assets
Prompt-to-image jobs can route through governed policies and return visible asset metadata.
Source-grounded visuals
Image and Vision Workflows
Move from simple generated images toward legal exhibits, infographics, diagrams, logos, and multimodal workflows.
Editing and diagrams
Future workflows include inpainting, upscaling, evidence diagrams, visual summaries, and branded graphics.
OCR and forms
Scanned documents, forms, screenshots, and exhibits can become source-anchored text and visual evidence.
Provenance metadata
Generated outputs should preserve route, prompt class, safety status, and asset lineage.
Safety gates
Unsafe or high-risk visual jobs block, escalate, or return safer alternatives.
In-chat display
Assets appear in the conversation so the workflow feels complete, not disconnected.