The collaboration software giant Atlassian, which has a market cap of roughly $47 billion, is launching its first major machine learning-based features for its flagship products.
Atlassian launched several capabilities across its products like Jira, Confluence, and Bitbucket on Wednesday that it calls "smarts," which can help users save time by making predictions on what they'll do.
To build these new features, the team tracked patterns about how its 150,000 cloud customers use its products. Those broad-strokes observations will be paired with a company's individual usage to create personalized suggestions for users.
"All these interactions allow us to have a data driven understanding of what you work on," founder of the Smarts and Machine Learning team at Atlassian, Shihab Hamid, told Business Insider. "By understanding these patterns, the machine learning gets smarter over time."
For example, Atlassian's products will suggest collaborators on projects or who should be brought into a conversation or code review — like recommending the top five most likely teammates to add to a project.
"That speeds up a common process that's repeated hundreds of times day-to-day," Hamid said. "We've built a slew of different intelligent experiences that help bring in the right person at the right time."
The new machine learning capabilities will also accelerate the process of handling IT issues by clustering together similar tickets based on summary, description, and categorization, eliminating the need for users to spend much time manually classifying IT tickets and learning from historical data. That way, users don't need to spend as much time classifying IT tickets.
"This really means teams can spend less time organizing tickets and more time solving the real problems at hand," Hamid said.
Finally, the company is rolling out smart search across its products that automatically surfaces what users frequently search for.
Atlassian started investing in building out the infrastructure to start the process of adding machine learning capabilities to its products in 2016, and it formed the Smarts team in 2017 to scale that up.
"In terms of investing in machine learning, we got the opportunity to accelerate teamwork," Hamid said. "If we think about products used in the market, it's all about tracking what you're working on. What we're beginning to see is it's not just data-in data-out. We're trying to optimize teamwork."
Got a tip? Contact this reporter via email at rmchan@bjinnox.com, Signal at 646.376.6106, Telegram at @rosaliechan, or Twitter DM at @rosaliechan17. (PR pitches by email only, please.) Other types of secure messaging available upon request.
