Listen to this article:
What's Actually Happening at GitLab?
The restructuring, called "GitLab Act 2", was internally announced on May 11, 2026 and confirmed during the quarterly earnings release on June 2. CEO Bill Staples described four fundamental changes in an official blog post:
1. Geographic consolidation. GitLab plans to exit up to 30% of the countries where it has small teams (specifically 22 countries). Local customers will continue to be served through the partner network.
2. Flattening the organization. In some departments, up to three layers of management will be removed. The goal is to bring leadership closer to the actual work and shorten decision-making processes.
3. Reorganization of development. Research and development teams will be broken down into roughly 60 smaller, autonomous units with full responsibility for their products. That's nearly double the current number of independent teams.
4. Internal deployment of AI agents. GitLab wants to be "customer zero" — the first user of its own platform. AI agents are to automate code reviews, approvals, and handoffs between teams. This will lead to "optimization of role counts."
Five Architectural Bets on an Agentic Future
In his letter, Staples describes five critical technology investments GitLab has staked its future on:
Infrastructure for machine scale. AI agents open dozens of merge requests simultaneously, run pipelines nonstop, and create commits at a pace no human team can match. Git itself wasn't designed for such load. GitLab is therefore rebuilding its foundational infrastructure — rewriting Git for the machine era, breaking the monolith into API-first services, and creating agent-specific APIs so that agents can act as first-class platform users.
Orchestration across the entire lifecycle. A single agent that writes code isn't enough for a company. It needs orchestration — coordination of agents across development phases, state management, context passing, conflict resolution, and policy enforcement. CI/CD pipelines, originally designed for human commit cadence, are being transformed into a runtime for agents.
Context as a key advantage. Most developer tools are converging toward similar code generation capabilities. What won't be commoditized is unique context — a connected data model spanning planning, code, reviews, security, deployment, and operations across all projects in a repository. GitLab is investing heavily in this layer.
Governance at the core. The more work agents take over, the more important control becomes: who can do what, what exactly happened and why, where sensitive data resides. GitLab is building identity, audit, and policies as fundamental platform services, not as an add-on product.
One platform, three modes. Companies will never switch to fully autonomous development overnight. GitLab is therefore building a single platform that supports human, assisted, and fully autonomous modes — with a shared data model and governance.
Source: BBN Times: GitLab Cuts 14% of Workforce
GitLab Duo Agent Platform: AI Agents Already in Action
The GitLab Duo Agent Platform, launched in January 2026, is the first concrete product of this strategy. It is an environment where developers can let AI agents perform complex tasks — from requirements analysis through code writing to security checks. The platform integrates with models from Anthropic (Claude Opus 4.8), AWS, and Google Cloud.
GitLab is also transitioning to a combined subscription and consumption-based billing model. Companies can buy both classic licenses and credits for work performed by AI agents. Other major market players are already following this model.
More detailed plans will be presented at the GitLab Transcend conference on June 10, 2026.
Financial Context: Growth and Investment Go Hand in Hand
Despite the layoffs, GitLab's numbers are strong. Revenue for the first quarter of fiscal year 2027 reached $264.2 million, a year-over-year growth of 23% and beating analyst expectations (which estimated approximately $254.6 million). Non-GAAP operating margin rose to 14% from the previous 12%, while GAAP net loss fell from $35.9 million to $5 million. The company generated a record free cash flow of $146.7 million.
The restructuring will cost $30–35 million (mainly in severance), with most of the savings to be reinvested into AI initiatives, not into margin increases. GitLab stock (GTLB) initially rose after the earnings announcement, but fell 5–8% in after-hours trading following the release of layoff details.
What Does This Mean for Czech Developers and Companies?
GitLab is one of the most widely used developer platforms in the Czech Republic — alongside GitHub, it is a key tool for code management, CI/CD, and DevOps. Czech companies that build their development workflows on GitLab should keep an eye on several things:
Availability and support. It hasn't yet been confirmed which 22 countries GitLab will exit. If the Czech Republic were among them (which is possible given the relatively small Czech team), customer support would shift to the partner network — which could mean no degradation if quality is maintained. GitLab explicitly states that "core services, support, and contractual commitments remain unchanged."
AI agents in Czech. The GitLab Duo Agent Platform currently works primarily in English, but with the advancement of multimodal and multilingual models (Claude, GPT), improvements in the Czech language environment can be expected. For Czech developers, this means that AI assistants will increasingly understand Czech-written documentation and code comments.
Pricing. Czech companies using GitLab Premium ($29/user/month) or Ultimate (custom pricing) should expect that agentic AI features will likely require an additional fee — whether in the form of consumption credits or a higher subscription tier. Detailed pricing for the European market hasn't been published yet.
A Broader Trend: Layoffs in the AI Era Aren't About Crisis
GitLab isn't alone. In 2026, we're seeing a new phenomenon: technology companies that are growing and profitable yet still laying off staff. It's not a response to a crisis, but a strategic redirection of resources into AI. The same path has been taken by Microsoft (which paused Copilot registrations due to infrastructure overload), Uber (which exhausted its AI budget in 4 months), and Salesforce (which turned Slackbot into an AI agent).
Bill Staples summed it up unequivocally: "Software will be built by machines, and managed by people." For developers, this means one thing — the ability to work with AI agents is becoming as important a skill as programming itself.
Will the GitLab layoffs affect Czech customers?
GitLab assures that core services, support, and contractual commitments remain unchanged. If the Czech Republic were among the 22 countries GitLab is exiting, direct support would transition to authorized partners. For the vast majority of customers, nothing will change — GitLab is a cloud service and its operations don't depend on the geography of its own offices.
What's the difference between GitLab Duo Agent Platform and GitHub Copilot?
GitHub Copilot is primarily a code-writing assistant in the editor — it completes lines, suggests functions, and helps with documentation. GitLab Duo Agent Platform goes a step further: it covers the entire software development lifecycle — from planning through code writing, reviews, security scanning, to deployment into production. It can coordinate multiple AI agents at once, automatically resolve merge conflicts, and enforce security policies. While Copilot enhances the work of an individual developer, GitLab aims at orchestrating entire development teams.
Does the GitLab Duo Agent Platform handle Czech?
The GitLab Duo Agent Platform currently works primarily in English, but uses models that understand Czech (such as Claude from Anthropic). For technical tasks like writing code, reviews, or managing CI/CD pipelines, Czech is not an obstacle. For multilingual documentation and comments, the quality of comprehension is continually improving.
