Requirement Generation
Intelligent Requirements can generate requirements from an initial scope description, business goal, problem to solve, supporting documentation, or reference images. The AI uses this input together with the selected knowledge bases to generate structured backlog items.
Generation happens level by level. Each level can use the items from the levels above it as a base, and child levels can also target a specific parent, such as generating stories for one epic.
The levels and naming of generated items depend on the selected provider: initiatives, epics, stories, and subtasks in Jira; epics, features, stories, and subtasks in Azure DevOps; the card types mapped in the workspace for BusinessMap. The goal remains the same: turn context into organized work items.
When to use
- Create a structured backlog from project context.
- Generate requirements that reflect documents and knowledge bases already prepared for the project.
- Expand an existing set of generated items as the demand evolves.
- Prepare the items that move on to review, refinement, saving, or publication.
Before you start
- Open the workspace that should contain the generated requirements.
- Select the knowledge bases that should guide generation.
- Confirm the provider model that defines the backlog structure.
- Prepare the scope, goal, problem, or supporting documentation.
- Prepare reference images when visual context helps, such as mockups, diagrams, or screenshots.
Step by step: generate requirements
- Open the workspace for the initiative.
- Confirm the selected project, knowledge bases, and provider.
- Add the initial scope, goal, problem, or supporting documentation, attaching reference images when visual context helps (up to 5 images, 5 MB each, in PNG, JPEG, or WebP).
- Adjust the generation options for the level, such as story type, persona, amount, and language.
- When generating a child level, pick the parent item when the new items should belong to a specific one.
- Start the generation flow.
- Review the generated backlog structure.
- Generate additional items when the demand needs to be expanded.
- Continue to review and refinement before saving or publishing items.
Generation options
| Option | Where | What it does |
|---|---|---|
| Story type | Stories | Switches between business-focused and technical stories, each with its own generation flow. |
| Persona | Stories | Defines who the story is written for, such as a product manager or an end user. |
| Technical profile | Subtasks | Defines the profile that will execute the tasks, such as a backend developer. |
| Amount | All levels | How many items to generate per run. |
| Language | All levels | Output language: Portuguese, English, or Spanish. |
| Reference images | All levels | Images used as visual context, such as mockups, diagrams, or screenshots. |
Expected result
You should have a structured set of generated requirements that can be reviewed, edited, saved to a knowledge base, or published to the selected provider.