The following page may contain information related to upcoming products, features and functionality. It is important to note that the information presented is for informational purposes only, so please do not rely on the information for purchasing or planning purposes. Just like with all projects, the items mentioned on the page are subject to change or delay, and the development, release, and timing of any products, features or functionality remain at the sole discretion of GitLab Inc.
| Stage | AI-powered |
| Group | Duo Chat |
| Maturity | Available |
| Content Last Reviewed | 2024-10-05 |
Thanks for visiting this category direction page on GitLab Duo Chat. This page belongs to the Duo Chat group of the AI-powered stage and is maintained by Torsten Linz (email).
This direction page is a work in progress, and everyone can contribute to our vision, architecture, and designs by commenting on the epic for the GitLab Duo Chat and its sub-epics and issues.
The recent advent of very capable large language models presents an opportunity to improve the way users interact with data. It is now possible to employ such AI models to let users interactively explore data via natural language.
The vision: GitLab Duo Chat is a conversational AI assistant that simplifies tasks and reducing context switches to empower users to achieve their DevSecOps goals faster and streamlined.
Initially, Chat will be limited in the types of queries it can answer. Long-term we intend for Chat to also be able to carry out tasks for the user.
Let's face it DevSecOps tasks are hard and complex. Understanding code that others have written, comprehending reported vulnerabilities, reading through tons of comments, understanding why a pipeline failed. And this is just the beginning, ultimately one wants to contribute to such code, fix the vulnerability, draw conclusions from the comments, and fix the pipeline.
Yes, such work is being done every day by developers, security professionals, product managers, and all the other personas that work in the DevSecOps space. However, they could do all these tasks much more efficiently and with less context switching with the help of AI.
GitLab Duo Chat aims to do just that and is a cornerstone in our FY25 Investment theme to make DevSecOps more efficient with AI.
Our Duo Chat strategy is to assist users with AI in ideation and creation tasks as well as in learning tasks across the entire Software Development Lifecycle (SDLC) to make them faster and more efficient.
We aim to employ the Chat for all use cases and workflows that can benefit from a conversational interaction between a user and an AI that is driven by a large language model (LLM). Typically, these are:
The Chat shall be context aware and ultimately have access to all the resources in GitLab that the user has access to. Initially, this context is limited to code files, code selections, the content of individual issues and epics, as well as the GitLab documentation.
To scale the context awareness and the supported use cases across the entire DevSecOps domain, Duo Chat aims to be a platform, that other GitLab teams and the wider community can contribute to. They are the experts for the use cases and workflows to accelerate.
Over the next 12 months, we will focus on the following prioritized themes, which represent broad categories of user needs that evolve as the market and our product mature. While these themes may extend beyond the 12-month horizon, we will continuously iterate and deliver improvements within each theme:
More details soon.
Details coming soon.
Details coming soon.
Details coming soon.
Competitors in the space include Android Studio Bot, GitHub Copilot Chat, Tabnine Chat, and Sourcegraph Cody AI Assistant.
GitLab Duo Chat aims to support across all there DevSecOps workflows. Therefore, all of GitLab's users are a target audience for the Chat. The initial focus is however on:
Duo Chat is availible as part of both Duo Pro and Duo Enterprise addons. Learn about how different Duo Chat features are tiered between the Duo addons.
Details coming soon.