Atlassian at the moment has unveiled a large suite of AI-driven updates at its annual Group occasion, signaling a significant shift towards an AI-native ecosystem. Central to the announcement is the enlargement of the Teamwork Graph and the evolution of Rovo, Atlassian’s flagship AI agent, that are designed to harness organizational context to ship unprecedented productiveness positive aspects.
The Teamwork Graph: A Unified Supply of Fact
The bedrock of Atlassian’s AI technique is the Teamwork Graph, a context layer that maps over 150 billion objects and relationships throughout a company’s tasks, targets, and groups.
In contrast to conventional AI instruments that depend on superficial searches and sometimes “stuff” context home windows with irrelevant knowledge, the Teamwork Graph permits AI to traverse deep relationships. Atlassian stories that this method makes AI solutions 44% extra correct whereas utilizing 48% fewer tokens.
To empower builders, Atlassian is releasing a Teamwork Graph CLI and MCP instruments. These permit clients to plug their organizational reminiscence into exterior AI builder instruments like Cursor or Claude, guaranteeing that custom-built apps profit from the identical wealthy context obtainable inside Jira and Confluence.
“We have already got over 150 billion objects and relationships mapped within the Teamwork Graph, and billions of these are altering each month,” Jamil Valliani, who runs the AI product crew at Atlassian, advised SD Occasions. “That could be a wealthy supply of fact for us and for our Rovo and different functions to make use of to ship more and more higher high quality outcomes and capabilities and instruments to our clients.”
Rovo Max: Pushing the Limits of AI Company
The highlight additionally fell on Rovo, Atlassian’s AI-powered software for looking out organizational knowledge, which has seen a 50% improve in utilization quarter-over-quarter, the corporate mentioned.
Atlassian at the moment launched Rovo Max, a brand new mode for fixing high-complexity issues. When activated, Rovo Max spins up a digital machine within the cloud to decompose duties, write code for evaluation, and take a look at itself autonomously. In a single demonstration, Rovo Max generated a professional-quality audio podcast summarizing a number of Confluence pages—a ability it was by no means explicitly taught, however discovered on the fly by researching finest practices.
Bridging the Hole Between Search and Motion
Past brokers, Atlassian is revolutionizing enterprise search. New connectors now hyperlink Rovo to over 50 third-party functions, together with SharePoint, Slack, and Salesforce. This integration permits Jira customers to reference Google Drive paperwork or Salesforce information instantly inside their workflow. “We are able to present high-quality search throughout all of these totally different functions,” Valliani mentioned.
In Jira, brokers have gotten first-class residents. Customers can now assign points to AI brokers as simply as to people. These brokers can seize plans in Jira, replace progress, and dialogue with crew members via feedback. Early knowledge reveals a 7x improve in agent-led automations, with firms like Mercedes-Benz already utilizing these instruments to enhance bug report high quality.
Valliani mentioned engaged on search is essential so clients can extra simply discover the paperwork that matter to them. “We ensure that we discover the fitting bits of knowledge for the client, each via graph and thru search, to energy the eventualities that matter most to them and use the very best knowledge.” That knowledge, he mentioned, isn’t just from Atlassian objects however consists of knowledge from the remainder of the group. “We’re capable of ship profoundly higher experiences and speed up workflows simply by having that functionality,” he mentioned.
Studio Updates
The Agent Constructing Studio has acquired vital funding to boost its capabilities past easy agent creation, now supporting automation and the event of entire new apps. The aim is to make AI constructing extra accessible, even for customers with out subtle technical information.
Valliani mentioned, “Customers can now merely use pure language prompts, resembling, ‘I have to automate X, Y, Z, in order that each day at 9am X, Y and Z occur.” The Studio routinely generates a plan, together with crucial brokers and automations, for evaluation. It has improved testing capabilities, permitting customers to dialogue with and take a look at the newly created agent in a panel. If there is a matter, customers can merely describe the issue, and Studio will right it, reasonably than requiring the person to replace the underlying immediate.
The Way forward for the SDLC and Developer Pleasure
For software program groups, Code Intelligence now gives semantic understanding of large codebases. Rovo can purpose over 20 years of code in minutes to establish type inconsistencies or safety vulnerabilities. Wanting forward, Atlassian teased “Rovo Dev,” which can quickly be able to autonomously submitting Pull Requests (PRs).
Reflecting on the enjoyment of improvement, Atlassian emphasised that these instruments will not be meant to exchange creativity however to take away the “tedious equipment” of standing stories and compliance. By delegating the nitty-gritty to AI, builders are free to give attention to fixing complicated issues and constructing merchandise that Valliani mentioned “make the client smile.”