GitLab’s Code Suggestions service is an AI-powered solution designed to optimize the software development process. It utilizes generative AI models to provide on-the-fly code recommendations, assisting developers in various languages and helping streamline workflows[1].
The feature was first introduced as a beta offering in GitLab 15.9 for GitLab SaaS (Software as a Service). After further enhancements, it was moved from GitLab Ultimate to GitLab Premium and then to GitLab Free in GitLab 16.0, before finally being enabled by default in GitLab 16.1. GitLab 16.1 also introduced support for Google Vertex AI Codey APIs, broadening its AI functionality and offering code suggestions in the GitLab WebIDE for all GitLab-hosted customers.
The Code Suggestions feature has been engineered to provide real-time coding suggestions in 13 different languages: C/C++, C#, Go, Java, JavaScript, Python, PHP, Ruby, Rust, Scala, Kotlin, and TypeScript. This feature is expected to help software developers boost their productivity, making everyday tasks like importing Python packages or completing functions more efficient. It can also aid new developers in exploring programming languages and eliminate the need for frequent web searches by experienced developers.
GitLab’s journey of integrating AI and ML (Machine Learning) into their DevSecOps platform is part of a broader plan to advance the capabilities of the software development process. By using AI-assisted code suggestions, developers are expected to be able to speed up tasks like coding, debugging, refactoring, documentation, and more, enhancing the software development lifecycle (SDLC). AI-assisted code suggestions use ML techniques to analyze code and recommend improvements. The aim is to help developers write better code faster by reducing errors, increasing effectiveness, and providing optimization advice.
As part of their commitment to extending the availability and utility of this feature, GitLab has made Code Suggestions available to all GitLab.com users for free while it is in Beta. The feature prioritizes privacy, ensuring that proprietary source code remains secure within GitLab’s enterprise cloud infrastructure and that such code is not used as training data.
In addition, GitLab expanded Code Suggestions to be available on Visual Studio through a Beta extension, allowing both GitLab SaaS and self-managed users to receive code suggestions while using the popular IDE. Once installed and configured, users can receive code suggestions as they type, enhancing their coding experience.
