Collab AI UX patterns
Collaboration patterns for shared sessions, presence, comments, and multi-user AI workflows.
Start here
Core patterns for collab UX.
5 patterns
Frequently asked questions
When do collab patterns matter for AI?
When multiple people share a thread, review outputs, or co-edit artifacts—the UI must show ownership, presence, and attribution.
How should shared session links handle permissions?
Offer view vs comment vs edit, expiring links, and clear indicators of who can see prompts and attachments. AI sessions often contain confidential uploads.
What is the difference between human-in-the-loop and comments?
Human-in-the-loop gates agent actions before they run. Inline comments annotate specific spans for human review—use both in approval-heavy workflows.
When is live presence necessary?
In shared canvases, paired programming, or workshop sessions where simultaneous edits collide. Async review can skip presence if comments and version history are strong.
How do smart diffs help AI collaboration?
They show what the model changed relative to a baseline so reviewers can accept/reject surgically—critical for docs, code, and contracts.
Which collab patterns support audit needs?
Shared sessions plus export and comment threads create a record of who approved AI output—pair with trust audit patterns for regulated teams.