Atlassian tested every confident AI take at scale and here is what actually worked.
Sherif Mansour oversees AI across Atlassian's entire portfolio plus 450 product managers. The company generates US$6.57 billion in revenue, employs 11,000 globally, and now has 5 million users regularly using AI features across 20+ applications. ANZ remains strategically important: Sydney hosts senior APAC commercial leadership including Jessica Gilroy, whose remit covers enterprise sales across ANZ, Southeast Asia, India, South Korea and Greater China.
They Almost Killed Chat
Two years ago, shipping Rovo chat to Jira and Confluence was controversial internally. The case against: chat is a terrible interface for structured work. The case for: nobody knew what users would actually do with AI.
Atlassian had to build a chat backend anyway to power AI features. Once it existed, they shipped it everywhere to learn fast. Millions now use it daily, even customers who also run ChatGPT and Claude. People use whatever AI is closest to where they are working.
What chat revealed: three patterns emerged from Confluence whiteboards. Users kept asking to group sticky notes into themes, so Atlassian built it as a button. They asked to turn brainstorms into Jira tickets, so that became a workflow. Complex cross-system queries stayed in chat. Every feature built for humans also had to be exposed as a tool for AI agents.
They Bolted AI Onto Jira Workflows
The standard advice is to rebuild from scratch. Atlassian did the opposite 2.5 years ago: they added one box to Jira's existing automation designer that let you insert a Rovo agent mid-workflow.
Customers chained agents together: ticket arrives, agent picks it up, marketing agent calls Canva to generate assets, social agent posts them. Enterprise teams already had workflows. They wanted AI inside those workflows, not a replacement for them.
Hiring Strategy Flipped
Atlassian historically hired senior engineers who could own complex distributed systems. That changed. Now they hire juniors who can ship fast with AI tooling. The calculation: velocity matters more than deep system knowledge when the tools handle complexity.
For ANZ sales teams covering Atlassian accounts, the product expansion matters. Cloud Collections are the growth lever, with service management, Rovo and AI capabilities driving upsell. The company competes directly with Microsoft, ServiceNow and Salesforce, while its own filings describe markets as fragmented and highly competitive. Worth watching: Atlassian is hiring solution sales executives specifically for ANZ and Southeast Asia enterprise accounts, indicating continued investment in consultative selling around complex implementations.
Subscription ARR sits at US$6.6 billion. Cloud revenue is growing faster than the overall business as customers expand usage across Jira, Confluence, Jira Service Management and AI capabilities. For account teams, that means cross-sell across 20+ applications with an installed base that is already using the core platform daily.