AI Assistants | Smart Data Generator Documentation
AI Assistants documentation.
Overview
AI Assistants bring the AI capabilities of Smart Data Generator directly into everyday Jira work.
Each assistant works where you already are — in the backlog or inside a work item — and always lets you review the result before anything is created or saved.
Available Assistants
| Assistant | Where to find it | What it does |
|---|---|---|
| Generate Sprints from Backlog | Board backlog → Generate Sprints | Suggests themed sprint names and creates up to 50 sprints in one action |
| Draft Comment | Work item → Content generator → Draft Comment | Drafts a reply based on the work item's summary, description and comments |
Context
Every assistant accepts context — a short instruction that tells the AI what you want to achieve, for example the theme of sprint names or the goal of a comment.
Context is optional in most cases. Each assistant's page describes when it is recommended or required, and how long it can be.
AI Usage on Evaluation Licenses
On evaluation licenses, each assistant has its own limit of 50 AI generations. The limits are counted separately for every assistant.
- Generations that fail (for example, because of a technical error) are not counted
- The number of remaining generations is shown in the assistant's dialog
If you need a higher limit while evaluating Smart Data Generator, contact us through the Customer Portal.
What Is Sent to the AI
Only the content you select in the assistant's dialog — for example the work item summary, description, selected comments and your context — is sent to the AI, and only when you start the generation.
Nothing is created or saved in Jira until you confirm the result.
Forge LLM
All AI Assistants use Forge LLM, the AI module built into the Atlassian Forge platform.
From the app's perspective, requests containing your data are processed within the Forge platform — Smart Data Generator does not send them to any external AI provider outside Forge.
Activity Stream Integration
Every assistant action is recorded in the Users Activity Stream, where you can review it and remove what was created.