What is Rovo?

Rovo is Atlassian’s AI offering that helps teams find personalized answers, automate tasks, and make smarter decisions using your organization’s connected context.

How Rovo uses the Teamwork Graph

Rovo is powered by Atlassian’s Teamwork Graph, our intelligent data layer that continuously maps relationships across your people, apps, projects, code, decisions, assets and more. This gives Rovo the ability to always answer and act on the latest information with accuracy and efficiency.

What are the Rovo apps?

Rovo is made up of three apps, along with a growing library of out-of-the-box agents:

  • Rovo Search: Search across Atlassian and your connected SaaS apps so you can quickly find the documents, data, or insights you need, no matter where they live. You can also use natural language to surface personalized results, like “Show me all Jira work items assigned to me”.

  • Rovo Chat: Chat with Rovo like you would a teammate. Ask questions, brainstorm ideas, and take action across your apps like creating project pages in Confluence or assigning Jira work items to your team directly in Chat.

  • Rovo Studio: Build your own AI agents, automations, and apps with Studio’s no-code builder. Use natural language and a library of Atlassian and 3P skills to describe your agent, or start from a number of agent templates and customize from there.

What are Rovo agents?

A number of Rovo agents exist out-of-the-box for you to start using today. Rovo agents are context-aware, designed to handle specific tasks, synthesize knowledge, and automate multi-step work across Atlassian and connected apps. Agents know how to act based on instructions, Atlassian and third-party skills, and organization knowledge that you specify.

Some examples of agents you get by default are Readiness checker, Release notes drafter, Brand voice crafter, OKR generator, Customer insights, and more.

To get the most out of Rovo, treat it as a collaborative partner rather than a magic solution. Guide it toward what you need, and continue to prompt to refine the output.