TLDR: GitLab is rolling out a new AI Agent Platform, including version 18.4, designed to significantly reduce operational ‘noise’ in DevSecOps workflows. These customizable AI agents, supported by a Knowledge Graph, aim to automate mundane tasks, improve security and compliance, and free up developers to focus on more strategic coding challenges, ultimately accelerating software delivery and enhancing quality.
GitLab, a leading DevSecOps platform, has announced a significant leap forward in its artificial intelligence integration with the deployment of advanced AI agents and a new Knowledge Graph. This initiative, highlighted in its latest release, version 18.4, is set to revolutionize how development, security, and operations teams manage their workflows by tackling the pervasive ‘noise’ that often hinders productivity. The core of this advancement is the GitLab Duo Agent Platform, now in public beta for Premium and Ultimate customers. This platform allows users to build, customize, and connect AI agents tailored to their specific work processes and organizational needs. These specialized agents are designed to automate a wide array of non-coding tasks that typically consume a developer’s time, enabling them to concentrate on critical business logic and creative problem-solving.
Emilio Salvador, GitLab’s Vice President of Strategy and Developer Relations, emphasized the company’s approach: “Our goal here is to meet developers where they are and make the entire developer experience as seamless as possible.” He further noted, “Basically, what it’s going to help developers [do] is focus on the things that matter most — solving business problems, working on business logic, and removing from the equation all those mundane tasks that developers don’t like to do.”
The AI agents are capable of automating tasks such as unit testing, checking security and compliance, modernizing Java code, automating code reviews, suggesting intelligent fixes, and optimizing testing and deployment decisions. For instance, a ‘Security Specialist’ agent can be trained on a company’s specific compliance rules to automatically flag issues in merge requests, while a ‘Documentation Pro’ agent can draft perfectly formatted guides from rough notes.
Beyond automation, the platform aims to create a ‘360-degree software development life cycle.’ Salvador explained that traditionally, software development has been a linear journey, but AI agents will close this loop by providing developers with real-time information from applications running in the cloud, detecting problems, and connecting them back to the developer.
GitLab’s Chief Information Security Officer, Josh Lemos, highlighted the role of AI in enhancing security operations. He explained that AI can “powerfully use AI to pull scraps of information together, get the right people to respond to the incident and pull the necessary data sources in order to meaningfully investigate the incident.” GitLab also leverages AI to identify anti-patterns in code, which are common but ineffective responses to problems. Lemos views AI as an “accelerant” for those with existing expertise, stressing the importance of human oversight, especially when dealing with less familiar programming languages. To address this, GitLab employs model routing on its backend, utilizing different Large Language Models (LLMs) for various programming languages.
Success metrics for this AI adoption include reductions in vulnerabilities and security anti-patterns, alongside faster resolution of issues. The platform’s extensibility is a key principle, allowing users to build their own agents and integrate them into existing workflows. The GitLab Duo Agent Platform also supports multi-agent workflows and offers agentic chat within IDEs, including JetBrains IDEs, enabling quick task delegation through slash commands like /explain, /tests, or /include.
Also Read:
- Aisera Advances Autonomous IT with Third-Generation Agentic AI Platform for Enhanced Enterprise Self-Service
- GroqCloud and IBM Watsonx Partner to Accelerate Enterprise AI Agent Deployment
This strategic deployment of AI agents, coupled with the contextual power of the new Knowledge Graph, positions GitLab to foster a future where human developers and specialized AI agents collaborate seamlessly, leading to increased productivity, higher software quality, and accelerated time-to-market.
