Rally Office Hours:
December 18, 2025

Summary

Product Announcement: AI-Driven Artifact Breakdown

A product speaker announced that a new feature is generally available for all users with the AI custom data module turned on.

Functionality

• The feature is an AI-driven breakdown of existing artifacts (like a feature) into children (like user stories or tasks).
• Users can break down any existing artifact with a single click.
• The Large Language Model (LLM) gets context from the artifact's description, acceptance criteria, children, and parent. Users do not need to copy/paste anything to the LLM.
• The process is currently in "silent mode," with a more conversational mode planned for the future.
• The AI uses an organization's existing templates to create the child items.

Benefits

• It will save a "dramatic amount of time".
• It is a collaboration between human and AI; the AI generates the initial set of children, and the product owner needs to review and refine them with the team.
• It provides consistency and well-structured work breakdowns.
• It allows for breakdown of all artifact levels, from strategy to task.

Process in Rally

• When breaking down an artifact (e.g., a feature), a user navigates to the Children page (or Tasks page for a user story).
• A new button, such as "create user stories with AI" or "create tasks with AI," will be visible.
• Automatically created children are in "silent mode," meaning they are created directly.
• Every AI-created artifact is given an "AI" prefix for the product owner to identify it, review it, and remove the prefix.

Q&A Session Topics

Topics discussed during the Q&A section included:

AI Feature - Going Backwards (Composition)

• An attendee asked if the roadmap included a feature to analyze the children of an artifact (e.g., user stories) and then update the parent's description or acceptance criteria.
• A panel member responded that this idea is "on the radar" but not currently in the backlog. The future idea is to be able to analyze changed children artifacts and offer suggestions to update the parent's description.

Query Box on Portfolio Kanban Board

• A panel member reminded attendees that the query box underneath the Portfolio Kanban board is going away.
• Teams should use the standard filtering side instead and save their configuration as a custom view, which can then be shared.
• The Portfolio Kanban board will also support custom states.

Custom Views Sharing Issue

• An attendee reported an issue when sharing a custom view with a list widget, where other users saw a default list instead of the administrator's configured list (columns, filters, group by settings).
• A panel member confirmed that the configured settings in edit mode should be persisted with the widget configuration and that the reported behavior sounded like a defect.
• The panel member suggested opening a support case.
• The panel member also noted that an administrator can restrict a user's ability to change filters, fields, or group by settings in view mode using the checkboxes below the view mode toolbar.

Capacity Planning for Kanban Teams

• An attendee whose team was switching from Scrum to Kanban asked for the best way to plan capacity for an entire PI (using a Release time box) since they could no longer use the Team Status view.
• A panel member suggested the Capacity Planning page as a good alternative.
• The Capacity Planning tool allows a user to:
  • Define an estimated capacity (e.g., 100 points) for a team during a PI.
  • Use the Capacity Forecaster for evidence-based capacity recommendations.
  • Select feature candidates for the PI.
  • View the feature's total story point rollup (sum of all associated stories) and compare it against the team's capacity.
  • Account for story points already completed within the features.

Forecasting Tools for Kanban

• A panel member suggested that the team could also use the Capacity Forecaster (found on the team board charts) to look at historical velocity and throughput. This tool helps predict how many work items can be done within a given time frame at various confidence levels.
• Another panel member suggested the Delivery Forecast tab on the team board charts. This tool flips the Monte Carlo simulation to estimate how long it will take to complete a designated amount of work based on historical performance.

Cycle Time Analysis for Kanban

• An attendee asked if there were similar forecasting tools for cycle time.
• Panel members suggested using the Scatter Plot:
  • It can be configured to show classic cycle time (e.g., "in progress to accepted") or use flow states for Kanban teams.
  • It shows the time it takes to complete work items at different confidence percentiles (e.g., 80th or 90th percentile).
• A panel member also suggested the Flowmetrics tab, which displays averages and trends for cycle time (or lead time, depending on the start state) and can be viewed by weeks or months.

AI Feature - Personas in User Stories

• An attendee reported that the AI-generated user stories default to the generic persona "as a user" and suggested a way to configure a table of real personas within a project.
• A panel member explained that the AI currently relies on the LLM to choose a persona from the context provided.
• A workaround was suggested: editing the user story template to include the phrase "as a [persona]" in angle brackets. The panel member claimed that in most cases, the LLM will replace this placeholder with a real persona based on the context.

Event News

Design Partnership

• There is a design partnership opportunity for the Insights product to work with users on measuring outcome and business value, including leading/early indicators.

Resources

• The product roadmap is available to watch anytime.
• The "Bridging the Gap" webinar is available on demand.

PI Planning

• An in-person PI Planning event will be held in February.
• Attendees can nominate themselves for the limited space to see how Rally uses the platform for PI planning. A panel member added that this is a "particularly good place to learn all about capacity planning" for both Kanban and non-Kanban teams.

Note: This summary was generated with the assistance of an artificial intelligence language model. While we strive for accuracy and quality, please note that the summary may not be entirely error-free. We recommend independently verifying the content and consulting with a product expert for specific advice or information. We do not assume any responsibility or liability for the use or interpretation of this content.

Ready to join the next Rally Office Hours session? Click the button below to sign up now!