Atlassian Engineers Demonstrate AI-Driven Software Development: From Tasks to Code Review Standards

For engineering leaders and developers seeking to deeply integrate AI agents into existing software engineering processes, a workshop published by Atlassian on October 11, 2026, offers a concrete practitioner perspective. Titled "The SDLC You Build Yourself," the session was led by Atlassian engineers Andy Wong, Andrei Bocan, and Shane Wolf. It focused on moving beyond simple in-editor code completion to workflows that start with a Jira item, draw on organizational knowledge, and produce changes for human review.

According to the official episode description, Wong demonstrated how to convert a design token standard into a reusable code review rule applied across selected repositories, rather than being enforced manually on every pull request. Bocan and Wolf explained the underlying context and orchestration, including selecting suitable backlog work, limiting cost and review volume, and asking people to clarify underspecified tickets before an agent proceeds. The description also notes that the discussion involved how the AI context engine handles code search, relationships between Jira and Confluence, and prioritizing verified sources amidst conflicting information.

Notably, network interruptions shifted parts of the session into discussions about sandbox verification, model selection, and deterministic checks. This resource is suitable for engineering teams evaluating how to leverage AI to connect current source code with historical reasoning while maintaining human authority. Readers are advised to look for specific timestamped segments on "Filtering noise from code review" and "Planning a feasible implementation" for more detailed operational insights.