Atlassian opens Teamwork Graph to third-party agents as Rovo evolves into autonomous executor

Atlassian Team 26 Pushes Teamwork Graph and Agentic Execution

Atlassian Corp. unveiled a comprehensive set of AI updates at its annual Team '26 conference in Anaheim, headlined by opening its Teamwork Graph to third-party agents and transforming Rovo AI from a helper into an agent capable of planning and executing multistep work autonomously2

2

. The Teamwork Graph, described as a living shared context layer connecting people, projects, documents and decisions across Atlassian and third-party tools, now contains more than 150 billion connections2

2

. The company is opening the graph through two new interfaces in open beta: a Teamwork Graph command-line interface for developers and Teamwork Graph tools delivered through Rovo's Model Context Protocol server[2](https://siliconangle.com/2026/05/06/atlassian-opens-teamwork-graph-p ushes-rovo-agentic-execution-team-26/).

The CLI, with more than 300 commands, lets coding agents such as Anthropic's Claude Code and Cursor query work and relationships across Atlassian products through a single interface rather than stitching together individual product application programming interfaces2

2

. Atlassian's own benchmarks show that grounding AI responses in Teamwork Graph data delivered 44% more accurate results while using 48% fewer tokens2

2

. More than 90% of enterprise cloud customers are now using Rovo AI agent, with customers performing more than 14 million Rovo-assisted actions in the past month2

2

.

New Tools Target Product Teams Under Structural Pressure

Atlassian shipped what reads as a prescription for product teams struggling with empowerment without structural support. Jira Product Discovery Enterprise reached general availability, and a new Feedback capability entered early access1

1

. The AI-powered Feedback capability captures and synthesizes customer signals directly into prioritization, addressing what Tanguy Crusson, Atlassian's Product Lead for Jira Product Discovery, calls the "second-hand information problem" where product teams rely on distilled one-liners that lack context1

1

.

Crusson explains the technical breakthrough that made Feedback possible: "The reason we didn't ship a feedback app before is we're not even sure that that was actually technically feasible, because it takes a lot of semantic analysis coupled with really good understanding of the domain model of the objects that we are talking about"1

1

. The promise is direct natural-language access to customer feedback, weighted and connected back to the customers giving it. Both Feedback and JPD Enterprise sit in a Product Collection alongside Roadway, a dynamic roadmapping app that helps teams rework priorities when goals shift1

1

.

Rovo Studio and Agentic Automations Reach General Availability

Rovo Studio, a no-code environment for building agents and automations grounded in the Teamwork Graph, reached general availability with built-in roles, approvals, versioning and audit controls2

2

. Agentic automations across Atlassian's platform are up sevenfold over the past six months2

2

. The company is adding a new reasoning mode called Max to Rovo Chat, available soon in early access, which breaks complex requests into multistep plans, executes them across connected tools, and loops users back in for review2

2

.

Agents in Jira are now generally available and can be assigned work items with full audit logging2

2

. A new Incident Command Center unifies incident detection, investigation and resolution with Rovo-assisted root-cause analysis, while Rovo Service offers autonomous or supervised Level 1 support2

2

. For engineering teams, Atlassian introduced Agent Experience for measuring how agents interact with codebases, AI Code Insights for tracking AI-generated code at the commit level, and AI Pulse for surfacing AI productivity signals to engineering managers2

2

.

Cisco Case Study Reveals Transform Don't Migrate Reality

Jason Andrews, VP of Engineering Operations at Cisco, presented a candid view of what enterprise AI adoption actually costs. Cisco collapsed 75 tools onto a single cloud platform, cut software spend by 54%, and generated an additional $5.3 million in annual savings from giving 10,000 users back 15 minutes a week3

3

. Andrews runs operations for a 20,000-engineer organization shipping more than 60 product lines and over $36 billion in annual revenue3

3

.

The transformation delivered a 3-5% productivity boost across Cisco's engineering teams, though Andrews caveated the measurement: "I didn't have a great way to measure it. But I went through and talked to the 50 engineering leaders, and the average response was three to 5%"3

3

. The timeline for Rovo adoption proved rougher than headline figures suggest. It took Cisco a year to get legal and compliance teams to approve use, with false starts when the tool was enabled then pulled back as teams worked through AI governance questions3

3

.

AI Governance Emerges as Critical Adoption Barrier

Kasia Wakarecy, VP Enterprise Applications at Pythian, a 450-person data and AI specialist, deployed Rovo to every single person in the company because it "extrapolated our existing privacy and security policies"5

5

. Pythian spent roughly five years getting its data governance foundation straight before plugging AI into core systems5

5

. Wakarecy warns that many organizations are doing this in reverse: "You cannot have good results unless you have your data shop in order"5

5

.

Connecting third-party applications to Rovo took less than a day at Pythian, but the real work had already been done on data clean-up and permission models5

5

. Internal surveys revealed a split adoption pattern: some employees use AI tools daily, while others have never touched them5

5

. Wakarecy describes the necessary approach as "kindergarten of technology," providing basic reassurance that users can touch the technology without breaking it5

5

.

Context Multiplied by Intelligence Defines Strategic Direction

CEO Mike Cannon-Brookes framed Atlassian's strategic direction around a formula: "Acceleration for your business is about context multiplied by intelligence. Intelligence is the engine, but context is the fuel"4

4

. In his view, 2026 marks the year raw model capability stopped being a differentiator, as organizations can "literally buy smarts by the token"4

4

. What cannot be purchased is institutional memory of failed launches, partial rollouts and incident threads that explain organizational decisions4

4

.

Internally, Atlassian is ingesting "multiple billions of objects every single week" into the Teamwork Graph, with the aim of propagating any change within 10 minutes4

4

. Customers are already running 5 million agent invocations a month on top of that context4

4

. Live demos during the keynote ran on production data, showing Rovo building briefings from 61 different sources spanning 20 years of customer interactions, and code search spanning 11 million files and 1.5 billion lines of code4

4

. The demonstrations worked, but as Cannon-Brookes acknowledged, "work will always be a little bit messy. That's where the human ingenuity actually lives"4

4

.