search
Cloud Blog – GitLab Duo Enterprise Sees the Light of Day
Gitlab

GitLab Duo Enterprise Sees the Light of Day

…and holds the potential to lighten the days ahead for many development teams all over the place. The roll-out came on the heels of GitLab’s recognition as an AI Code Assistance Leader on the 2024 Gartner® Magic Quadrant™. Making the debut chart in the category, GitLab stood out in both vision and execution domains, also bringing home the titles for less-novel albeit just as significant ‘AI-Powered Capabilities Across the SDLC’ and ‘Transparent and Privacy-First AI.’

With LLM-enabled personal coding assistants gaining momentum, scaling was just a matter of time. Well, the time has come, and GitLab has officially extended the award-winning experience to entire teams and organizations. In this article, we meet GitLab Duo Enterprise, explore its differentiators, and look at its impact on business.

GitLab Duo Enterprise vs. GitLab Duo Pro

Let’s start by examining how GitLab Duo Enterprise stacks up against its less-advanced counterpart, GitLab Duo Pro, features-wise.

FunctionalityGitLab Duo EnterpriseGitLab Duo Pro
An ability to grant and revoke user permissions
Code production
Code finalization
Cross-IDE availability
Supported programming languages25+15
Code walkthrough
Code refactoring
Test generation
Summarized discussions
Summarized merge requests
Summarized code reviews
Vulnerability assessment
Vulnerability remediation
Root cause analysis
AI impact analytics
Model deployment in a self-hosted settingComing soon
Model personalizationComing soon
PricingUS$39 per user per month under 1Y commitmentUS$19 per user per month under 1Y commitment

Please also keep in mind that the Pro edition is available to both Premium– and Ultimate-tier customers, while GitLab Duo Enterprise is Ultimate-only. As for the features that are a work in progress—namely, self-hosted deployment and personalization—some extra terms and fees may be applicable.

GitLab Duo Enterprise Features

This section delves into each individual functionality that’s only available as part of GitLab Duo Enterprise.

Discussion Summary (built on top of Anthropic Claude 3.5 Sonnet)

In a highly collaborative environment, an issue thread can turn into a long conversation in no time. Naturally, that often makes it challenging to keep stakeholders in the loop. To battle this bottleneck, GitLab Duo Enterprise has introduced Discussion Summary. This feature distills lengthy comment threads into short yet informative summaries and provides a quick overview of key action points and decisions made. The time they spent sifting through endless comments stakeholders can now devote to making meaningful contributions to ongoing discussions.

Merge Request Summary (built on top of Vertex AI Codey text-bison)

When reviewing code changes, developers often find proposed updates scattered across multiple files. This results in a limited ability to quickly grasp the overall impact. Merge Request Summary makes things easier by auto-generating a broad summary of these updates. If a developer puts this summary into the merge request description, they provide a clear and to-the-point overview of the suggested modifications, which, in turn, encourages informed discussions and decision-making that follows.

Code Review Summary (built on top of Vertex AI Codey text-bison)

When it comes to code reviews, feedback—however detailed—is still spread across lines and lines of code. That’s why GitLab Duo Enterprise comes with Code Review Summary: the goal is to simplify the review process and ensure clear communication. How that’s achieved is through an automatic, compact overview of all comments left by reviewers. With this summary, handover becomes super-fast so that neither reviewers nor developers miss a single potential improvement ever again.

Vulnerability Explanation (built on top of Anthropic Claude 3 Haiku)

Without proactivity in vulnerability management, it’s virtually impossible to maintain application security. Running scheduled scans on your code or CI/CD pipelines, GitLab Duo Enterprise surfaces potential risks and presents them in a centralized report. When you select a specific vulnerability, you’ll gain access to its precise location in your code and the scanner that flagged it. Vulnerability Explanation then breaks down the nature of the vulnerability and where it originates from, outlines potential ways to exploit it, and most importantly, provides clear guidance on how to come up with swift remediation.

Vulnerability Resolution (built on top of Anthropic Claude 3.5 Sonnet)

Once the vulnerability is located and understood, GitLab Duo Enterprise goes a step further and instantly produces a merge request to address the issue. With remediation automated, Vulnerability Resolution brings down the time and effort needed to respond to security risks a great deal so developers can concentrate on their primary tasks knowing that their app is protected at all times.

Root Cause Analysis (built on top of Anthropic Claude 3.5 Sonnet)

Sometimes, CI/CD pipelines fail, and troubleshooting ends up a time-consuming effort. Root Cause Analysis speeds up the process as it automatically analyzes logs and returns a complete, human-readable explanation of the failure along with recommendations on how to tackle it. As a result, GitLab Duo Enterprise reduces troubleshooting overhead and saves developers valuable time, letting them focus on innovation and deliver software faster.

AI Impact Dashboard

The AI Impact Dashboard is a go-to tool for performing a health check on your artificial intelligence initiatives. It allows you to track the effectiveness of AI across your organization based on SDLC metrics such as DORA, visualize the positive outcomes, compare the performance of AI-enabled teams versus traditional ones, and monitor the overall progress of AI adoption.

Any Tangible Results Yet?

Cube, a software development company based out of the Netherlands, delivers high-quality and secure mobile app, website, and e-commerce solutions across the energy, real estate, food delivery, and wellness sectors. To meet ever-growing demands of its customers, the company was looking for an AI-powered SDLC platform that could make their development processes more complete.

Following a game-changing experience with the Ultimate plan, Cube selected GitLab Duo Enterprise for its advanced AI features and capabilities. It offered a suite of tools that immediately ‘clicked’ with the development team, including code suggestions, test generation, and summarizations.

So what’s changed with GitLab Duo Enterprise on board? One, it is way easier to carry out code reviews; two, there is not much room for an element of routine anymore. In the pre-AI era, developers had to go through long comment threads, whereas now they can locate critical issues on the spot and start addressing them right away. Bottom line? The code quality has improved, and that is a direct result of better feedback visibility and prioritization.

As there is no such thing as ‘too much security,’ the company has also made use of automated scanners to review all new code and enforce pre-approved security rules. This degree of standardization ensures that vulnerabilities are consistently addressed, regardless of the developer’s experience or expertise.

The Cube’s leadership is particularly excited about security-focused AI capabilities that detect and alert developers about code vulnerabilities before they even find their way into merge requests. In that regard, the company seeks to adopt merge request summaries, vulnerability explanations, and issue description generation across the entire SDLC.

The results so far are pretty impressive:

  • A 50% improvement in lead times.
  • A 50% acceleration of vulnerability detection.
  • 40 hours of dev time saved weekly.

About Us

Cloudfresh is a GitLab Professional Services Partner and GitLab Select Partner recognized globally. In terms of GitLab Duo Enterprise, we render the following services:

  • Consulting.
  • Implementation.
  • Value discovery.
  • Core principles explanation.
  • Q&A sessions.
  • Training sessions.
With all the AI fever, we get it. If you’d like to put GitLab Duo to a test before committing, just let us know by pressing the button below. Request a free trial →
CTA Image