Atlassian has integrated AI capabilities into more than 20 of its applications, with over 5 million users regularly engaging with these features, the company revealed at SaaStr. Sherif Mansour, Atlassian's Head of AI, shared insights from his 17 years at the company and his role overseeing product management for 450 product managers. The AI integration spans both legacy and newer apps built during the generative AI era.
Two years ago, Atlassian faced internal debate over adding a chat interface, called Rovo chat, to Jira, Confluence, and other apps. Critics argued chat was a poor user experience for many tasks, such as filling out forms, while proponents saw value in providing an open-ended interface to observe user behavior with AI. The decision to include chat was ultimately driven by the need to build a chat backend to support AI features across the portfolio, enabling the company to deploy chat in all apps and gather user data.
Atlassian’s approach highlights common challenges for B2B software companies integrating AI. The company tested multiple strategies, including whether to bolt AI onto existing products or rebuild from scratch, and debated the role of chat as the primary interface versus embedding AI into real features. Their experience with a large, diverse product suite and millions of users offers a data-driven perspective on AI adoption in enterprise software, contrasting with more theoretical industry claims.
Sherif Mansour emphasized that Atlassian’s AI journey involved balancing user experience with technical infrastructure, ultimately shipping AI features across all 20+ apps. This deployment supports a broad user base of over 5 million, demonstrating significant scale for AI in established SaaS products, according to saastr.com.