For a considerable period, Atlassian has provided collaboration solutions, frequently preferred by developers and IT professionals, with well-established applications like Jira for support requests, Confluence for work organization, and Bitbucket for managing development outputs. However, the platform previously lacked an integrated machine learning capability to enhance user efficiency both within and across its various applications.

This situation evolved today with Atlassian’s announcement of the development and release of several tools leveraging a new machine learning layer, known as Atlassian Smarts. Notably, unlike companies such as Salesforce, with its Einstein layer, or Adobe, with Sensei, Atlassian deliberately avoided adopting a branded marketing name, opting instead to emphasize the technology’s inherent capabilities.

Shihab Hamid, the founder of Atlassian’s Smarts and Machine Learning Team and a 14-year veteran of the company, explained that the decision to forgo a marketing name was intentional. “Our primary focus is on the user experience, and rather than emphasizing the technology through branding, we are dedicated to streamlining teamwork,” Hamid stated to TechCrunch.

Hamid clarified that the machine learning layer’s purpose is to simplify the complexities associated with coordinating individuals and information throughout the platform.

“Even straightforward tasks, such as identifying the appropriate individual or locating a specific document, can present challenges, potentially hindering productivity and diverting time from more valuable, creative endeavors. Teamwork is inherently complex, and collaboration can be difficult. These are fundamentally human issues without simple solutions,” he noted.

He explained that Atlassian is addressing these challenges through machine learning, aiming to expedite repetitive and time-consuming processes. Similar to the approaches taken by Adobe and Salesforce, Atlassian has constructed an underlying layer of intelligent functionality, which can be implemented across its platform to deliver machine learning-driven features where they are most beneficial.

“We have invested in integrating this functionality directly into the Atlassian platform to connect IT and development teams and unify their work. Consequently, core Atlassian products like Jira and Confluence are built upon this shared platform and benefit from its consistent functionality. By establishing this common predictive capability at the platform level, we can effectively distribute intelligence and leverage the data collected across our products,” Hamid said.

The initial implementations of this vision are now available. First, Atlassian is introducing an intelligent search tool designed to help users locate content across Atlassian applications more quickly by understanding their roles and work patterns. “By recognizing where users work and their areas of focus, we can proactively provide access to relevant documents and accelerate workflows,” he said.

The second component focuses on enhancing collaboration and assembling teams with the most suitable personnel for specific tasks. A new feature called predictive user mentions assists Jira and Confluence users in identifying the right individuals for a project.

“We have integrated this intelligence into the Atlassian platform, leveraging our knowledge of user work patterns and collaborative relationships to predict who should be included in discussions,” Hamid explained.

Lastly, the company unveiled a tool specifically for Jira users that groups together similar support requests, potentially leading to faster resolutions compared to handling them individually.

“We are soon releasing a feature within Jira Service Desk that enables users to cluster comparable tickets, accelerating IT workflows. This is achieved through machine learning techniques that calculate ticket similarity based on summaries, descriptions, and other factors,” he said.

This development was facilitated by the company’s prior transition from primarily on-premises solutions to the cloud, which provided the flexibility to create new tools that operate across the entire platform.

Today’s announcements represent the initial phase of Atlassian’s plans to introduce a continuous stream of new machine learning-powered features to the platform in the months and years ahead.