Atlassian’s Rovo, an AI-Driven Search Engine for Enterprises, Is Now Generally Available

Screenshot of Atlassian's Rovo agent builder. Image credit: Atlassian

Five months ago, Atlassian introduced an AI agent equivalent of an enterprise search engine that sourced information from its family of apps and also third-party providers. Powered by Atlassian Intelligence, Rovo is designed to help workers better discover knowledge, learn from it and take action. Today, it’s generally available to all Atlassian customers.

Rovo has three parts, all of which can operate independently or, when combined, establish a workflow with the potential to improve discovery and decision-making. The first is Rovo Find, which indexes content from internal and third-party tools to put it in one place. The second is Rovo Learn, a tool to facilitate education, or, in other words, “a tailored company glossary for your organization that updates in real-time.” Lastly, there are Rovo Agents, specialized bots organizations can use in workflows to address time-consuming tasks, complete projects and more.

Atlassian boasts Rovo has saved early users an average of one to two hours a week, enabled 75 percent to accelerate their work production, and helped 80 percent find the correct information at the right time. “Feedback has been pretty exciting so far,” Jamil Valliani, the company’s vice president and head of AI product, tells me in an interview. “The early signals are really positive to us, and we’re continuing to use feedback to go and drive improvements.”

Chief executive Mike Cannon-Brookes agrees, saying in a statement, “Rovo is one of the most ambitious and important products we’ve ever built…Whether you’re a dev team building the next big thing, a marketing team creating your next big campaign, or a service agent helping your customers solve a really big problem, AI is going to transform the way your work gets done.”

What’s New With Rovo?

In addition to Rovo being open to the public, Atlassian is highlighting multiple updates it has released over the past few months to make its enterprise search engine more capable.

Rovo Find

The enterprise search engine is the part of Rovo that consolidates all information into a single place. In May, it had access to a limited number of subscription apps such as Google Drive, Microsoft Sharepoint, GitHub, Slack, Microsoft Teams and Figma. Valliani says over the next six months, the company hopes to expand that list to 80 “of the most popular SaaS applications.”

Social signals seen in Atlassian Rovo Find search results. Image credit: Atlassian
Social signals seen in Atlassian Rovo Find search results. Image credit: Atlassian

Atlassian has also enhanced the search results screen to showcase social signals, displaying relevant content, such as when a colleague commented or edited a document that might be relevant to you. Using the above example, Valliani explains that Brooke’s conversation is shared because Rovo recognizes that you are colleagues, sending a signal that it’s probably relevant. That information is also used to boost content rankings.

Another new feature activated last week is bookmarks, which isn’t the same as if you want to bookmark a webpage in your browser. Instead, this is helpful for administrators to surface important content that’s frequently searched for, such as “Where do I get a pay stub? or “Where do I file expense reports?” Within Atlassian, admins will specify the query and the URL for the answer and designate that as a bookmark. Users will see that at the top of their results page — it’s like an ad at the top of a Google search page, but placed for free.

And speaking of Google search results, Rovo Find’s design closely resembles that of the leading search engine. Specialized cards appear depending on what you’re querying, one of which is knowledge cards. If Rovo knows an answer, it’ll display the card featuring insights about the projects, goals, who’s working on it, and more. But in the past few months, Atlassian has added a new card type dedicated to helping you better know about your teammates. When shown, it will display not only their name and contact information but what they’re working on and who else they’re working with, so you feel more connected and prepared to collaborate with them — they aren’t just a digital avatar you see on a Slack thread.

Lastly, Atlassian is introducing a data center connector for customers who use the company’s data center products. The idea is to make Rovo Find accessible to those who cannot use Atlassian’s cloud offering for whatever reason (e.g., regulatory). The search feature will initially support Confluence starting in early 2025 before adding JIRA afterward.

As Valliani describes it, because AI is generally cloud-powered, it becomes inaccessible to those whose data is locked away behind a walled garden. “But what we’re trying to do is say, if you can work with us on a connector basis, we can stream the data we need to do the AI work, especially in search, from the data center to the cloud, hold it security on our side, and then power a certain set of removal capabilities, especially in search and chat, to go make that possible for you, so you can benefit from those things without taking everything out into the cloud, which often they need more time, other hoops to be jumped through.”

Rovo Learn

An example of Atlassian's Rovo Chat in action in Confluence. Image credit: Atlassian
An example of Atlassian’s Rovo Chat in action in Confluence. Image credit: Atlassian

This component’s signature is a chat feature. This is the interface you can use to ask Rovo about the status of a marketing campaign, receive feedback or resolve issues. “It’s not just any chat,” Valliani says. “It really is the go-to expert on your organization that can learn from it in any context without having to make you switch between different tools or information down, and you can do all the same things you can do on ChatGPT on top of that.”

Initially, this chatbot was only accessible within Atlassian’s platform, though the company’s head of Atlassian Intelligence, Sherif Mansour, disclosed that an off-platform browser extension would be coming. It’s now here.

An example of the Atlassian Rovo Chat browser extension in action. Image credit: Atlassian
An example of the Atlassian Rovo Chat browser extension in action. Image credit: Atlassian

If you’re using non-Atlassian software supported by Rovo, and your administrator has enabled access to that third-party app, you can use the Rovo Chat browser extension for assistance. “It can start understanding the context even when you’re not inside Atlassian and answer questions to you right then and there,” Valliani claims. “So, for example, you’re reading a Google Doc that’s a project proposal. It’ll be able to go and see…terms you may not understand. And we can help you get those definitions from Rovo, and then you can ask Rovo Chat questions about that document without ever having to leave that page.”

Valliani asserts that no personal documents or data will be used to augment Atlassian’s teamwork graph. Rovo must be in a signed-in state to do whatever it does. I’m told that the only data pushed to the teamwork graph and into the work context are through the connectors established by a company’s Atlassian administrators.

He highlights that Rovo Chat is one of Rovo’s highest-retained features. Customers “when they start using chat, they’re like ‘Oh, my God, this is so good!’ And they use it at phenomenally high rates, which has been really excellent for us. We’re continuing to promote that quality as well as discoverability—the ability to go and actually launch chat from different places.” As it turns out, the Rovo Chat browser extension wasn’t something the company was working on six months ago.

Rovo Agents

What’s an AI tool without agents? “We view them as these teammates you can pull into your workflow that save you hours or days of work while also helping you get better outcomes,” Valliani remarks. “They can be added and referenced as if they’re normal teammates, making it easy for customers to get to know and interact with them, but they only enter upon invitation.”

Initially, a limited number of agents were available, including Backlog Buddy. This bot helped teams with their work backlog, reducing the amount of manual human work hours spent completing the task. Atlassian disclosed at the time that at least 15 companies were building agents for Rovo.

A list of some of the AI agents created for Atlassian's Rovo. Image credit: Atlassian
A list of some of the AI agents created for Atlassian’s Rovo. Image credit: Atlassian

At launch, 20 out-of-the-box agents are now available. Anyone signed up for Rovo can access agents such as Decision Director, Rovo Expert, User Manual Writer, OKR Generator, Comms Crafter, Global Translator, and Release Note Drafter.

Valliani expects customers to want to create their own agents because “there are so many company-specific workflows.” As with other agent platforms, Atlassian customers won’t need extensive technical knowledge to program their agents to handle company onboarding for new hires, enforcing brand and style guidelines, or anything else their team thinks up.

Showing the Value

“I think a lot of folks have…opened their eyes to agents in the last six months, which is great,” Valliani says. “People are affirming this idea that your AI will become your teammate. And, we’re the teamwork company. We think that’s a very compatible vision for us and a direction we love. We’re focused on helping our customers apply these agents into their workplace. We’ve proven now, through our beta program and with our Rovo release, that we have the right tools. We can actually create and are already using agents inside our own company.”

He references a customer 360 agent an Atlassian employee spun up in a matter of days that allowed executives to ask how many JIRA licenses a particular customer company has, when was the last time they met with their CEO, or about who the CEO is.

Valliani intimated that Atlassian’s focus over the next few months would be on sales, finding ways to convince its customers to create agents that work for them, embracing the technology, and fully adopting Rovo.

When I raised a concern some company executives might have about not finding value in AI, something raised by Salesforce Chief Executive Marc Benioff last month, Valliani responded, “With any new technology, there’s a period of time where we have to help everyday people learn how to use it.” He explained that it’s both a user-side challenge, where people need to acclimate to AI, and a technology issue, where the interface must be made simple and clearly show how it can solve users’ problems.

He compares AI to the early days of personal computers when the interfaces and applications needed to evolve to become more user-friendly and valuable. The Atlassian executive expects a similar evolution in making AI tools more intuitive and impactful. He’s optimistic about the future of AI in the workspace but stresses that it’s critical to focus on solving real problems for customers and making the technology easy to adopt.

Leave a Reply

Discover more from Ken Yeung

Subscribe now to keep reading and get access to the full archive.

Continue reading