What They Did
GitLab is an all-in-one DevOps platform — source control, CI/CD, security scanning, and project planning in a single application. Built by one of the most prominent all-remote companies in tech, it grew to roughly 2,500 employees spread across dozens of countries and became the main open-core alternative to GitHub for enterprises that wanted to self-host their software supply chain.
How LLMs Killed Them
GitLab's business is priced per developer seat, and the developer seat is exactly what agentic coding tools are compressing. On June 2, 2026, the company cut about 350 people — roughly 14% of staff — and announced it was exiting 22 countries, flattening management layers, and partnering with an unnamed AI lab to rebuild its platform for agent-scale workloads. CEO Bill Staples framed it as "GitLab Act 2," a "generational rebuild" of core infrastructure for what he estimated as a 100x increase in traffic, arguing that agentic workloads were "pushing competitors to the brink." The tell is the timing: GitLab announced the cuts on the same day it reported Q1 revenue of $264M, up 23% year over year. Growth was not the problem. The problem was that the money had to move from headcount to infrastructure fast enough to serve software agents rather than the humans who used to write the code.
Timeline
- 2023–2025: GitLab shipped Duo, its AI assistant suite, as coding agents began writing, reviewing, and merging code at scale.
- June 2, 2026: Reported Q1 revenue of $264.2M, up 23% YoY — and announced 350 layoffs (~14% of staff) the same day.
- June 2026: Announced exit from 22 countries, flatter management, and a "generational rebuild" of core infrastructure for agentic workloads.
- 2026: Expects $30–35M in restructuring charges as it redirects spend from people to AI infrastructure.
By the Numbers
- 350 jobs cut — about 14% of the workforce
- Exiting 22 countries as the all-remote company contracts
- Cuts announced the same day as 23% YoY revenue growth
- $30–35M in restructuring charges
- CEO Bill Staples: rebuilding for 100x agent-driven traffic growth