DevOps Engineer

We are currently seeking a DevOps Engineer [ML Infrastructure] to join our team.

Starting task: Bring up a production-ready Kubernetes cluster on Hetzner for ML training and backtesting workloads. There are some artifacts from a previous team — you'll need to review them, decide what's worth keeping, and rebuild the rest. This isn't an out-of-the-box job; we need someone who knows how to run Kubernetes on bare metal, not just on managed cloud offerings.

Core responsibilities:

• Deploy and maintain a Kubernetes cluster on Hetzner for ML workloads (model training, backtests), including GPU/CPU resource management, scheduling, and storage

• Build a full observability stack: Prometheus, Grafana, centralized logging, alerting

• Automate server provisioning and configuration with Ansible / Terraform

• Set up CI/CD pipelines for image builds and deployment of ML components

What we're looking for:

• Hands-on experience setting up Kubernetes on bare metal (Hetzner, private DCs, or similar) — not just managed clusters like EKS/GKE

• Solid command of Prometheus, Grafana, and a logging stack (Loki / ELK / equivalent)

• Production experience with both Ansible and Terraform

• CI/CD pipeline design for container images (GitLab CI / GitHub Actions / ArgoCD / similar)

• Understanding of ML infrastructure specifics: GPU nodes, job queues, large datasets, long-running jobs

• Ability to read and assess legacy infrastructure code and make pragmatic decisions about what to reuse vs. rewrite

• Independence — this is a sole-DevOps role on the project

Nice to have (bonus scope):

• Experience with trading, low-latency, or data-heavy systems

• VPN setup, network policies, secrets management (Vault)

• Familiarity with ML tools

What we offer:

• Remote cooperation

• Modern and complex projects using ML technologies

• The possibility of professional and financial growth

• A single focused project — no context-switching across unrelated work