Contextual Conversations
Chat in Product Graph isn’t just a generic AI assistant—it’s deeply integrated with your product documentation. The AI understands:- Your existing artifacts and requirements
- Project context and goals
- Team terminology and conventions
- Historical decisions and their rationale
Key Capabilities
Requirements Refinement
Discuss and refine requirements through natural conversation. The AI helps
identify gaps and ambiguities.
Idea Exploration
Brainstorm features, explore alternatives, and think through implications
before committing to documentation.
Documentation Generation
Generate drafts of specifications, user stories, and acceptance criteria
from your conversation.
Quick Answers
Ask questions about your existing documentation and get instant, contextual
answers.
Adding Context
Enrich your conversations by bringing in additional context beyond what’s already in your product documentation.Attachments
Upload files directly to your chat to give the AI more context:- Images: Screenshots, diagrams, wireframes, and mockups
- PDFs: Research documents, specifications, and reference materials
Support for text files including markdown, plain text, and DOCX is coming
soon.
Artifact Mentions
Reference other artifacts in your messages using mentions. Type@ to search and select any artifact in your product. The AI will pull in that artifact’s content to inform its response, making it easy to ask questions that span multiple documents or compare requirements across features.
Document Editing from Chat
When you ask the AI to edit your document through chat, changes are applied as tracked revisions—never directly overwriting your content. The AI proposes changes; you review and approve.Creating Child Artifacts
Beyond editing the current document, Chat can create new artifacts in your product hierarchy:- Propose before creating: The AI always proposes what it plans to create and waits for your confirmation
- Batch creation: Multiple related artifacts (like several requirements) can be created in parallel
- Automatic linking: New artifacts are properly linked to their parents in the hierarchy
The AI will never create artifacts without first explaining what it intends to
create and receiving your approval.
Example Prompts
Requirements refinement
- “What edge cases should we consider for the user authentication flow?”
- “Can you help me write acceptance criteria for this user story?”
- “What’s missing from this requirements document?”
Documentation generation
- “Generate a technical specification from these requirements”
- “Create a user story for the shopping cart feature”
- “Write API documentation for the endpoints we discussed”
Document editing
- “Add a section about security considerations”
- “Make the overview more concise”
- “Rewrite this in a more technical tone”
- “Break this long paragraph into bullet points”
Context questions
- “What decisions have we made about the payment system?”
- “Show me all requirements related to user onboarding”
- “What was the reasoning behind using GraphQL?”