Continuously evaluate agent performance against real MRs and pipeline runs, and iterate on prompts, tool definitions, and orchestration logic to improve reliability
Collaborate with platform and release engineering teams to ensure agent-driven changes meet the same quality bar as human-authored changes
Maintain and extend internal APIs/SDKs that let other teams build their own agents on top of the shared pipeline infrastructure
Contribute to open-source agent tooling and represent the company in relevant upstream communities
Write clear technical documentation on agent capabilities, limitations, and pipeline architecture
What We're Looking For
Required
3+ years of experience building production software, with hands-on experience building LLM-powered agents or tool-using systems
Strong proficiency in Python, plus experience with GitLab CI/CD (.gitlab-ci.yml, runners, pipelines, webhooks, GitLab API)
Practical experience with agent frameworks or patterns (function/tool calling, ReAct-style planning, multi-agent orchestration) — whether via LangChain, LangGraph, custom orchestration, or similar
Experience integrating LLMs into developer tooling (e.g., automated code review, CI failure triage, test generation)
Solid understanding of CI/CD concepts: build systems, artifact management, deployment gating, and pipeline security
Familiarity with Linux-based development environments and version control workflows
Strong engineering fundamentals: testing, code review, observability, and safe rollout practices for automated systems
Preferred
Experience building guardrails for autonomous systems (permissioning, sandboxing, human-in-the-loop approval flows)
Familiarity with self-hosted GitLab administration or GitLab Duo / other AI DevOps integrations
Experience with vector databases or retrieval-augmented generation (RAG) for codebase-aware agents
Background in release engineering or build systems for a Linux distribution or large open-source project
Experience with containerization and orchestration (Docker, Kubernetes) for running agent workloads in CI
Contributions to well-known open-source Linux or AI agent projects
Nice to Have
Experience with evaluation frameworks for agent reliability and regression testing
Familiarity with Rust or Go for building performant pipeline tooling
Public speaking, blogging, or conference talks on AI agents, DevOps automation, or the Linux community
Why Join Us
Work on an OS used by millions of developers and enterprises worldwide
Meaningful open-source contributions as part of your day-to-day work
Small, senior team with real ownership over architecture and roadmap
Competitive salary, equity, and benefits
Flexible remote-friendly culture
