Tabnine is proud to announce our participation as a launch partner for Atlassian’s soon-to-be-released AI integrations, dubbed Atlassian Rovo, previewed today at Atlassian Team 2024. Once available, these new integrations will allow Tabnine’s customers to leverage the complete body of information — the content and code that represents deep institutional knowledge — that’s captured across Atlassian’s suite of products. Additionally, Tabnine’s first-to-market AI coding assistant will be embedded within the Atlassian product suite as an AI agent to provide our full scope of capabilities anywhere asoftware development team works.
We’re excited about what the future holds. Atlassian opening their products to be better accessed by AI-powered software development tools, along with Tabnine’s deep expertise as the originator in the coding assistance space, will further accelerate and simplify our joint customers’ software development effort. And as an embedded agent, Atlassian customers will benefit from Tabnine’s ability to generate code, fixes, tests, and documentation with full awareness of their codebase and software development practices.
We’re also working on expanding our integrations with Atlassian products, and a few days ago provided a glimpse of what’s coming soon: an AI agent that can take simple specifications stored in Atlassian Jira and create a fully functioning application.
But you don’t need to wait for tomorrow: Tabnine and Atlassian are already working together to make engineering teams more successful. Read on to learn more.
Tabnine has a long-established relationship with Atlassian. Tabnine is not only an Atlassian Ventures-backed company but is also a long-time Atlassian customer.
Over the years, we’ve built deep integrations with Atlassian’s products. Tabnine customers already use our connections to Atlassian’s products to get more personalized AI assistance using the context derived from accessing the information in Atlassian products.
By design, the LLMs that are integral to every AI coding assistant are universal. Even though they are trained on vast amounts of data and contain billions of parameters, they’re not aware of the specific code and distinctive patterns of an individual organization. As a result, their recommendations, even though accurate, are quite generic and not tailored to the needs of a developer. To increase the effectiveness of AI coding assistants, it’s imperative to provide contextual awareness to the LLMs so that they can understand the subtle nuances that make a developer and organization unique.
Retrieval-augmented generation (RAG) is a technique that’s widely used in the industry to provide context to coding assistants. RAG not only reduces LLM hallucinations but also helps to overcome the inherent limitations of training data. Tabnine uses RAG and other advanced techniques to leverage your codebase and knowledge to make our AI coding assistant align with how you work and provide more accurate and highly personalized results.
Here’s how you can use Tabnine with Atlassian’s products today:
Context through local code awareness: Tabnine integrates with Atlassian Bitbucket (alongside other leading Git-based repositories like GitHub and GitLab). Start by checking out the branch from Bitbucket from the Jira issue you’ve been assigned and open the project in your IDE. Tabnine then accesses the locally available data in your IDE, including variable types used near the completion point in the code, comments you’ve added, open files you’ve interacted with, imported packages and libraries, and open projects. Tabnine automatically identifies the relevant information and uses it as context to provide personalized results. We’ve seen that personalized AI recommendations based on awareness of a developer’s IDE are accepted 40% more often than AI suggestions generated without these integrations.
Connection to your Bitbucket repository for global code awareness: Often the most valuable context for an engineering team typically exists beyond just what’s available in the developer’s IDE. This is especially true for enterprises where teams of hundreds of developers work cross-functionally to build software. Tabnine administrators can connect Tabnine to their organization’s Bitbucket repositories (alongside any other Git-based tools like GitHub or GitLab) to significantly increase the context and get highly personalized results when completing code, explaining code, creating tests, writing documentation, and more. In addition, this connection allows a developer to use plain language queries to better understand an existing codebase, to find code that serves specific functions or leverages specific APIs, and to identify code that can be reused on their project. This capability is currently in Private Preview for Tabnine Enterprise customers. Check out the video below to see the integration between Tabnine and Bitbucket in action:
Customization of AI models: Building on the personalization of the AI assistant through context and connection, Tabnine offers model customization to further enrich the capability and quality of the output. We use your codebase stored in Bitbucket to fine-tune the proprietary models that we custom-built for software development teams. Leveraging your codebase to extend Tabnine’s existing models results in higher performance in common software development tasks and dramatically improves the quality of code generated for companies that use less common programming languages or frameworks.
Want to see how Atlassian + Tabnine can help your engineering team be more productive? Reach out to schedule a chat with one of our product experts. If you happen to be in Las Vegas at Team 2024 this week, stop by Booth #68 to learn more about these capabilities. Or to try it out for yourself right now, sign up for Tabnine Pro today — it’s free for 90 days.
Explore recent posts
//Explore recent posts
//
2 -min read
7 -min read
4 -min read
More Announcements
//2 -min read
4 -min read
6 -min read