The way development teams work is changing. Engineers now juggle dozens of tools, repositories and communication channels at the same time. Every time they open another tab to track down a Jira issue, read a Confluence page, or search for a design file in a third‑party app, they lose focus. Over an entire week those small interruptions add up. Atlassian’s recent announcement of Rovo Chat for Bitbucket speaks to this problem. Rovo Chat brings generative AI into Bitbucket so developers can ask natural language questions and get answers right inside their code view. Instead of toggling between tabs, they have an ever‑present teammate that surfaces context, summarizes conversations and even helps them understand unfamiliar code. This blog unpacks the key points of our recent presentation about Rovo Chat, why it matters for developer productivity and how you can start using it in your own workflow.
Information overload: In Atlassian’s 2025 State of Developer Experience report, over half of developers said they lose more than ten hours every week just searching for information, onboarding to new code or switching between apps. That is more than a full workday wasted on tasks that add no direct value.
Fragmented tools: Teams often rely on Jira for issues, Confluence for documentation, Figma for design and other tools for monitoring and feedback. Each tool stores a piece of the puzzle. When developers need an answer, they may not know where to look.
Context switching: Every time a developer leaves their code editor to open a browser tab, their brain has to switch contexts. Research suggests employees spend around 1.8 hours a day searching for information. Those lost minutes compound, slowing progress and increasing frustration.
Inefficient onboarding: New hires spend days or weeks piecing together what a repository does by sifting through pull requests, reading comments and tracing dependencies. When experienced engineers change teams, they face the same challenge.
Time‑consuming code reviews: Reviewing pull requests often requires scrolling through long threads, following comment chains and hunting for related Jira tickets or design docs. Without clear summaries, it is easy to miss important details.
Rovo Chat addresses these pain points by connecting the dots across Atlassian products and third‑party apps. It surfaces knowledge when and where developers need it, reducing the mental overhead that comes from juggling multiple tools.
Rovo Chat is part of Atlassian’s larger Rovo platform. Rovo was introduced at the Team ’24 conference as an AI‑powered solution for finding, understanding and acting on data scattered across Atlassian and third‑party apps. Built on Atlassian’s Teamwork Graph a data fabric that maps relationships between people, projects and content Rovo can interpret the intent behind queries and return contextual answers. Unlike traditional search, which is confined to a single product, Rovo understands how a Jira ticket relates to a pull request or how a Confluence page connects to a Figma file. Rovo comprises three core components: Search, Chat and Studio. Search lets users query information across Jira, Confluence and connected apps through a unified interface. Studio provides a workspace where teams can build AI agents that act on their behalf, such as responding to common questions or automating repetitive tasks. Rovo Chat sits between these two. It allows you to converse with your data using natural language. You might ask about the status of a sprint goal, the purpose of a repository or the dependencies in a file. Rovo Chat will analyze information from Jira, Confluence and other connected sources to generate an answer. Within Bitbucket, Rovo Chat appears as a button in the top navigation bar. When clicked, it opens a chat interface on the right side of the screen. Because it lives inside Bitbucket, it has access to the pull request or file you are viewing and can respond to prompts based on that context. Rovo is not just a chatbot bolted onto Bitbucket; it is integrated into the product so that your conversations are anchored in the code you are reviewing. When you ask a question about a function or file, Rovo analyzes the code diff or repository metadata to craft its response.
Summarizing pull requests: Reviewing long pull requests can be tedious. Rovo Chat can summarize the code changes or highlight key comments. By asking “Summarize the code changes in this pull request,” you receive a concise overview that tells you what changed and why. This helps reviewers focus on the most important parts.
Onboarding to new repositories: When you start work on an unfamiliar codebase, you can ask Rovo questions such as “What’s the purpose of this repository?” or “What are the key APIs and dependencies in this file?”. Rovo will analyze the repository and provide context so you can ramp up faster.
Linking Jira and Confluence context: Developers often need to know which Jira issue a code change relates to or where a design file lives. Instead of hunting through tickets, you can ask Rovo to show you the project page for a feature or point you to the design document. Rovo retrieves the relevant Jira issue or Confluence page and presents it in the chat.
Summarizing documentation: Confluence pages can be long and detailed. Rovo can read a page and extract the key points. This saves time when you need a quick overview before diving into the details.
Explaining code and answering questions: Developers frequently jump between files to trace a function’s behavior. When you encounter unfamiliar code, you can ask Rovo to explain it. Rovo will provide an explanation based on the surrounding code and documentation. You can also ask general coding questions, such as “What does getByLabelText do?”, without leaving Bitbucket.
Team insights: Engineering managers can use Rovo to see which pull requests their teams are working on and who has reviewed them. This visibility helps managers stay on top of progress without micromanaging.
These use cases illustrate how Rovo Chat reduces friction in everyday development. It operates within the flow of work, meeting developers where they are rather than forcing them to adopt a separate tool.
Rovo Chat is currently in beta. Atlassian began a phased rollout in late September 2025 and plans to make it available to all Bitbucket Cloud users by September 30th 2025. If you see a Rovo Chat button in your Bitbucket navigation bar, you already have access. If not, you should gain access soon. To enable Rovo Chat, your workspace administrator must connect your Bitbucket workspace to Jira and enable AI features in both Jira and Bitbucket. Indexing your data can take a day or two, so give Rovo time to process your repositories, issues and pages. Once indexing finishes, the Rovo Chat button will appear for users, and you can start asking questions. If your organization uses Bitbucket Data Center, you can try Rovo by signing up for a free trial of Bitbucket Cloud Premium. As of the beta release, Rovo Chat works with Bitbucket Standard and Premium plans. It is integrated with the broader Rovo platform, which supports more than fifty third‑party apps such as Figma, Sentry and Dropbox. This means you can connect design files, error logs, document storage and other services to provide richer answers. Rovo continues to evolve, so keep an eye on Atlassian’s announcements for additional integrations.
Pull requests and repositories: Rovo indexes pull requests and repositories in Bitbucket. When you ask a question, it can analyze the contents of a pull request or the file you are viewing and include that context in its response.
Active code and diffs: Rovo can only read code and diffs that are currently open on your screen. It does not scan your entire codebase unless you ask it to in the future. This scoped indexing helps protect sensitive information.
Connected Atlassian apps: Rovo connects to Jira, Confluence and Jira Service Management, indexing issues, pages and requests. It understands relationships between these items through the Teamwork Graph, so it can map a pull request to a Jira story or a design page.
Third‑party data sources: More than fifty external apps can be connected to Rovo. These include design tools, monitoring services and file storage platforms. When you connect a third‑party app, Rovo indexes metadata from that service so it can surface relevant information in your chat.
Privacy and permissions: Atlassian emphasizes that Rovo respects existing permissions. It only retrieves information you already have access to and follows the privacy policies outlined in Atlassian’s documentation. When indexing external apps, you must authorize Rovo to access them, and you can revoke access at any time.
Understanding what Rovo indexes helps teams manage expectations. The current beta release focuses on active content within Bitbucket and connected sources. In future releases, Atlassian plans to extend indexing to full repositories and other workflows.
Rovo Chat’s beta is only the beginning. Atlassian has shared a glimpse of features planned for the general availability release. One major enhancement is full code context: Rovo will eventually index your entire codebase, not just the files you have open. This means you will be able to ask Rovo questions like “Has anyone already written a function that handles OAuth authentication?” and get an answer based on code across your organization. Such a feature could dramatically reduce duplication and encourage reuse.
Another upcoming feature is CI and CD insights. Atlassian plans to let you ask Rovo about build status and pipeline results without scrolling through log files. You might ask, “What caused the last build to fail?” or “Which commits slowed down our test suite?” and get a summarized explanation. These insights will help developers troubleshoot issues faster and understand the health of their pipelines at a glance.
Beyond these specific enhancements, the roadmap includes deeper integration with the broader Rovo platform. As Rovo Search and Studio mature, Rovo Chat will be able to trigger actions through AI agents, such as creating Jira issues, sending meeting notes or recommending next steps based on conversation context. Future updates could also include customizable prompts and private model tuning to align answers with organizational standards.
Be specific with your queries: Rovo works best when you ask focused questions. Instead of a general request like “Tell me about this project,” ask “What are the main components of the billing service and who maintains them?” A clear prompt helps Rovo deliver precise answers and reduces follow‑up queries.
Use conversational language: You do not need to structure your queries like commands. Rovo is designed to understand natural language. Feel free to phrase questions as you would ask a colleague. For example, “Could you summarize this pull request and highlight any changes to authentication?”
Connect relevant apps: To get the most out of Rovo, connect the apps your team relies on. If your design team works in Figma and your operations team uses Sentry, connecting those services allows Rovo to surface design files or error logs in response to questions. Make sure your workspace administrator reviews permissions and grants access only to necessary data sources.
Review and refine: Rovo’s answers are generated by AI models trained on your organization’s data. While they aim to be accurate, always review the suggestions before acting. If an answer seems off or incomplete, provide feedback through the interface. Atlassian encourages beta users to share feedback to help improve the tool.
Following these practices will help your team adopt Rovo smoothly and avoid frustration. As with any AI tool, thoughtful use and clear communication are key.
While this blog focuses on Bitbucket, it is important to see Rovo Chat as part of a broader strategy to unify work across Atlassian tools. Rovo Search enables users to find information across Jira, Confluence and other connected apps through one interface. This unified search breaks down product silos: a query about a new feature can return Jira stories, Confluence specs and design files from one place. Rovo Chat, meanwhile, allows conversation around that information. For example, a product manager might ask about the status of a marketing campaign and get data pulled from multiple sources. Rovo Studio extends the platform by allowing teams to build AI agents. These agents perform tasks like responding to customer questions, onboarding new team members or clearing out a backlog. By leveraging agents, organizations can automate repetitive processes and free humans to focus on more strategic work. The combination of Search, Chat and Studio means Rovo is not just a chatbot but a comprehensive AI layer that sits across Atlassian’s cloud platform. In addition, Rovo leverages the Teamwork Graph to maintain context about relationships between people, projects and content. This graph ensures that answers are specific to your team’s workflows. It understands who owns a service, how that service relates to a Jira epic and which documents support it. This contextual awareness is what sets Rovo apart from generic chatbots or search tools.
Rovo Chat addresses developer pain points: It reduces time spent searching for information, minimizes context switching and accelerates onboarding. By surfacing summaries and connecting data across tools, it helps teams stay focused.
Integration with Bitbucket keeps you in flow: Rovo appears as a chat panel within Bitbucket, accessing the pull request or file you are viewing. This integration provides in‑context answers without leaving your code.
Getting started is straightforward: Connect your Bitbucket workspace to Jira, enable AI features and wait for indexing to complete. Once ready, the Rovo Chat button appears and you can start using it.
The roadmap promises deeper capabilities: Full code indexing and CI and CD insights are on the way. As Rovo evolves, expect more powerful features and tighter integration with AI agents and third‑party apps.
Ready to explore Rovo Chat further? Atlassian is actively seeking feedback during the beta period. By trying the tool now, you can shape its development and help tailor it to real‑world workflows.