Highlights:

  • GitHub Copilot Enterprise acts as a developer companion, allowing them to pose questions about both public and private code.
  • It faces competition from Tabnine Chat, a generative AI-based code completion tool.

GitHub, owned by Microsoft Corp., launched GitHub Copilot Enterprise, an advanced AI-powered coding assistant tailored to company codebases and processes.

Named GitHub Copilot Enterprise, the tool is now widely accessible. It functions as a developer companion, enabling them to inquire about both public and private code, acquaint themselves with unfamiliar codebases, and establish uniformity within engineering teams, according to the company.

Additional advantages involve the ability to tailor chat conversations to the organization’s code repositories, ensuring that responses are consistently derived from its internal knowledge base. Also available is a web search function powered by Bing that is in beta.

GitHub specified that the enterprise-grade copilot tool is exclusively accessible to subscribers of GitHub Enterprise Cloud, with a cost of USD 39 per user per month.

The initial GitHub Copilot for individual developers was introduced in 2022, followed by a business-oriented edition. GitHub Copilot Enterprise, unveiled in November, represents a more sophisticated iteration. Similar to its forerunners, it seamlessly integrates into the user’s development environment, offering real-time coding suggestions and comments during their work. However, it surpasses them with additional features exclusive to this advanced version. GitHub also highlighted the inclusion of enterprise-grade security in GitHub Copilot Enterprise.

As an example, it facilitates semantic search based on repositories, simplifying the process for developers to explore code repositories. Furthermore, it can examine request differentials, aiding developers in comprehending proposed code changes more effectively. Another noteworthy feature is its capability to tap into internal knowledge bases, enhancing the precision of coding suggestions.

Over the next few weeks, GitHub Copilot Enterprise will introduce support for finely tuned models. This will empower users to train the chatbot using their own codebases, enhancing its predictive capabilities and responses.

Competing with GitHub Copilot Enterprise is Tabnine Chat, a code completion tool based on generative AI. While the pricing is comparable, Tabnine Ltd.’s offering is more cost-effective. This is because GitHub users need a GitHub Enterprise subscription to access the predictive capabilities of GitHub Copilot Enterprise, making Tabnine Chat a more budget-friendly alternative.

GitHub’s CEO, Thomas Dohmke, expressed that the preceding versions of Copilot have ushered in a new era of software development, resulting in evident improvements in developer productivity and satisfaction. A six-month study involving 500 developers that was conducted by Accenture Plc is cited to support the company’s position that claims regarding increased developer productivity are thoroughly warranted. 90% of those developers surveyed in the study stated that Copilot’s assistance improved the quality of their code.

Moreover, the study asserted that 88% of the code suggestions from Copilot were ultimately adopted, resulting in a 50% enhancement in the number of builds produced.

GitHub is promoting Copilot as a fundamental component of the developer experience. However, the technology has raised several concerns, such as its inclination to recommend fee-based tools, its reliance on public codebases, and the potential risk of introducing bugs and vulnerabilities through its code suggestions.