Extending GitHub Copilot: insights from Atlassian, Elastic, and Split

Introduction - Welcome to webinar rel="noopener noreferrer" target="_blank">(00:00:00)

• GitHub Copilot extensibility is introduced.
• Copilot can be extended to support custom languages, frameworks, and tools.
• Atlassian uses Copilot to generate boilerplate code, documentation, and test cases.
• Copilot helps Atlassian engineers to be more productive and focus on more complex tasks.
• Elastic uses Copilot to generate code snippets, documentation, and test cases.
• Copilot helps Elastic engineers to be more productive and focus on more complex tasks.
• Elastic is exploring ways to use Copilot to improve the user experience of their products.
• Split uses Copilot to generate code snippets, documentation, and test cases.
• Copilot helps Split engineers to be more productive and focus on more complex tasks.
• Split is exploring ways to use Copilot to improve the developer experience of their products.
• Questions from the audience are answered.

Atlassian - Demo + Q&A rel="noopener noreferrer" target="_blank">(00:00:03)

• GitHub Copilot extensibility allows integrators to extend GitHub Copilot chat, enabling specific interactions tailored to development needs.
• Extensions can respond to user requests via existing AI solutions, a proprietary language model, or even Copilot itself.
• Extensions can make requests against the GitHub API and enhance the developer experience within the IDE or github.com.
• Partners like Sentry, LambdaTest, and LaunchDarkly have developed extensions showcasing various use cases.
• Atlassian's extension provides context from Jira and Confluence without leaving the editor, reducing context switching and improving developer efficiency.
• The extension is powered by Atlassian's Roo, a large knowledge model that offers unified search and connectors to third-party tools like GitHub.
• The Atlassian GitHub Copilot agent provides additional context from Jira and Confluence, resulting in more accurate and descriptive responses to user queries.
• GitHub Copilot can bring context from various Atlassian products into the IDE, making it easier for developers to write code.
• The integration with Atlassian products can surface historical data from Jira and Confluence, helping developers work more efficiently.
• GitHub Copilot can reduce interruptions and mundane tasks, allowing developers to focus on writing code, learning, and collaborating.

Elastic - Demo + Q&A rel="noopener noreferrer" target="_blank">(00:25:33)

• Elastic is a search company that provides search engines, custom search applications, observability tools, security tools, and advanced AI capabilities.
• Elastic's observability features include collecting logs, metrics, APM data, synthetic data, user experience data, and universal profiling of the environment.
• Elastic's AI assistant helps operators understand complex issues by providing insights, error messages, culprits, and stack traces.
• GitHub Copilot is an AI assistant that can help developers write code by providing tools and functions to find alerts, provide context, and run queries.
• Elastic has released an AI assistant API that allows developers to integrate AI functionality into their own applications.
• Elastic is currently working on a VS Code extension that will provide AI assistance within the development environment.
• Elastic AI assistant is currently accessible through Kibana, but the goal is to integrate it into the GitHub Copilot extension, allowing users to interact with it directly from their IDE or GitHub Copilot chat.

Split 00:19:10 - Demo + Q&A rel="noopener noreferrer" target="_blank">(00:55:23)

• GitHub Copilot is an extension for Split that allows developers to interact with their Split setup directly within their IDE.
• Copilot can provide information about feature flags, including their status, traffic distribution, and definitions.
• Copilot can also generate code to implement functions needed to use the Split SDK.
• Copilot currently has read-only capabilities but aims to expand into write operations in the future.
• Copilot is exploring the ability to manage configuration data directly within the IDE.
• To join the GitHub Copilot partner program or for inquiries about the extension, contact support@split.io.
• Other AI tools that integrate with IDEs include Atlassian's project and team-specific knowledge from Jira and Confluence, and Elastic's AI tooling.