Atlassian argues that most developer time (84%) is spent outside of coding, in activities like clarifying requirements, searching for context, and attending meetings. To address this, they advocate for an AI operating model that connects workflows across the entire SDLC rather than applying AI only in the code editor. Using their own Confluence engineering team as a case study, they show how Rovo, Jira, Confluence, Loom, and the Teamwork Graph work together to unify design, planning, implementation, and post-launch feedback loops. Reported outcomes include 500,000 meeting hours saved, 100+ hours per engineer per year, 45% improvement in PR cycle time, and 50% faster release rollouts. A practical playbook covers starting with focused pilots, using a hub-and-spoke governance model, and involving IT and security early.
Table of contents
From AI features to an AI operating modelHow Atlassian’s engineering teams embed AI across the SDLCMeasurable impact for engineering teamsA pragmatic playbook for AI adoption at scaleMoving from fragmented tools to a unified, AI-native systemTeamwork Collection as a connected system of workSort: