Senior DevOps Engineer
Build and maintain reproducible Linux infrastructure for software development, AI research, and production environments. Support developers and researchers with system, networking, build, and GPU issues.
Responsabilidades
- Design and maintain reproducible NixOS infrastructure, reusable configurations, packages, and modules.
- Troubleshoot Linux systems and performance across CPU, memory, storage, and kernel settings.
- Manage system services and investigate issues with tools such as strace and tcpdump.
- Configure networks, routing, firewalls, VPNs, access controls, and secrets management.
- Develop infrastructure tools using Nix, Bash, and Python or Rust.
- Automate configuration builds, testing, evaluation, and deployment through CI/CD.
- Deploy and scale local LLM inference services and infrastructure for agentic workflows.
- Create consistent development environments with required SDKs, toolchains, and environment variables.
- Support colleagues with environment, build, GPU driver, and networking issues.
- Write documentation and troubleshooting guides, and help teammates build Linux and Nix skills.
Requisitos
- Strong Linux administration skills and understanding of kernel and userland environments.
- Experience with performance tuning, systemd, journald, and low-level troubleshooting.
- Hands-on Layer 2/3 networking experience, including VLANs, bonding/LACP, and custom routing.
- Experience with firewalls, VPNs, secrets management, and authentication integrations such as SSH, OIDC, or LDAP.
- Ability to build maintainable tools using Bash and Python or Rust.
- Experience with CI/CD, advanced Git workflows, and code reviews.
- Ability to deploy and operate local inference services and supporting AI infrastructure.
- Ability to troubleshoot GPU drivers, CUDA, native dependencies, and build failures.
- Clear communication, patience, a service-oriented approach, and commitment to reproducibility and documentation.
Se valora
- Experience with NixOS, custom packages, modules, and Flakes.
- Familiarity with non-standard build systems, patches, and native C/C++ dependencies.
- Experience with multi-node deployment tools such as Colmena, Morph, NixOps, or Deploy-rs.
- Familiarity with NVIDIA Blackwell systems, including DGX B200 or B300.