Atlassian Corp. Plc. said today it’s bringing generative artificial intelligence agents and services online across its flagship collaborative product software line so its customers can get the benefits of conversational access to their organization’s data and quick insights.

In April, the company announced the first set of services as part of Atlassian Intelligence, launched in beta mode at the time, and developed as part of Atlassian’s own in-house AI models. It also uses OpenAI’s GPT-4, which is one of the large language models that underlies its popular ChatGPT chatbot.

“Many of our products help teams plan and track projects, or deliver a service, whether they be technical or non-technical teams,” Sherif Mansour, head of Atlassian Intelligence, told SiliconANGLE in an interview. “And we think that AI is going to help our customers achieve great things and accelerate their productivity with the different types of teamwork that we serve our customers in.”

The company’s AI offerings are backed up by more than 20 years of public and customer data that can be brought to bear for individuals as virtual agents and large organizations that can tap into their troves of information in ways that they have never been able to before.

Today, generative AI is now available in Jira, the company’s project and issue tracking software. That will permit users to generate customer responses quickly that are factual or empathetic within the editor. It’s also generally available in the Confluence team collaboration platform, allowing teams to quickly get up to speed about projects by summarizing documents and asking questions of the AI.

The same AI service can use organization and domain-wide information to help new users integrate themselves into the culture so that they don’t feel left out of important information such as jargon terms. It’s much better than an employee handbook wiki or a search, explained Mansour, because a virtual agent can handle questions and answers.

“One great example is your classic company glossary like all these, a lot of our big customers have a ton of content in Jira, Confluence and Trello, with a lot of internal jargon and acronyms that could be project names or system names,” said Mansour. “One of the things we’re delivering is a definitions capability where you can highlight any content in Confluence, and say, explain what is ‘Project Veritas.’”

From there the AI will respond with conversational language, like talking to another person who knows the subject. If that doesn’t explain it well enough, the user can ask further questions and drill down further. This opens up a lot of opportunities to explain otherwise esoteric company jargon and get new employees up to speed faster with onboarding.

Right now, this new capability is integrated into Confluence in beta mode, but soon it will be able to pull content from other Atlassian products.

Already 10% of Atlassian customers have been putting the new AI services to use, Mansour said. For example, Domino’s Pizza Inc. has put it to use to manage its post-incident reports. The powerful summarization capabilities allow management to get high-level takeaways quickly, saving hours reading reports by eliminating the need to dive into the details to get insights.

“In my day-to-day work, Atlassian Intelligence acts as a virtual teammate that summarizes five-page PIR reports into five-sentence recaps for me, so I can get up to speed quickly ahead of monthly review meetings,” said Matthias Hansen, group chief technology officer at Domino’s. “It’s already proven its worth by improving productivity across our product teams.”

Accessing information from vast troves of enterprise data is equally important for companies. Atlassian announced the launch of natural language prompts for Jira Query Language, which will allow users to dig deeper into Jira issues, by combining the ease of basic search and also allow users to build complex queries with JQL and discover tracked issues in the system quickly.

With natural language searches for SQL, nontechnical teams can access insights and ask questions through Atlassian Analytics to explore insights and visualize important business metrics. That means a manager could ask a question about business trends, customer support trends or team health.

Ordinarily, a data scientist would have to write a SQL query to dive into company data to pull this sort of information and then format it. Now a simple conversational query can prompt the AI to generate the underlying SQL code for reports.

Knowledge stored in Confluence can be accessed through AI agents with questions and answers as well by employees with the full context of whom they are, said Mansour. For example, it would be aware of what projects they’re working on or what locations they work in.

“You could ask it about an HR policy: ‘What’s our leave policy for annual leave?’ and it’ll pick up your location and tailor the response,” he said. “So if you’re in Sydney, it would it would say, this is the policy for staff in Australia versus those that are in other locations.”

This new capability is currently in beta mode for Confluence. A similar question-and-answer capability is coming in beta mode to Compass, the company’s developer experience platform that helps track and discover software components. Using the same AI agent technology, users can ask questions about their software stack and reduce time locating information about their tools and microservices.

Images: Atlassian