Senior DevOps / MLOps

VICTORY IT seeks a Senior DevOps / MLOps in Yekaterinburg (hybrid, full-time) to build a distributed compute platform for LLMs and ML. Salary $5,000–10,000 net/month. Kubernetes/Nomad, GPU clusters, Ray, CI/CD.

VICTORY IT is looking for a Senior DevOps / MLOps to join the large language models and machine learning team. We will build a distributed computing platform that unites CPU/GPU and other compute resources into a single infrastructure for ML tasks. What you will work on: • Design and build from scratch a distributed compute platform for ~1000 machines. • Develop Kubernetes / Nomad infrastructure for running and managing ML workloads. • Work with GPU clusters and different compute resources — GPU, CPU, MPS. • Develop distributed computing on Ray and infrastructure for distributed training. • Automate task placement and launch, resource management, and utilization. • Handle scaling, performance, fault tolerance, CI/CD, and monitoring. • Work closely with the ML team on infrastructure for training and running neural networks. Note: this is not classic DevOps for supporting existing infrastructure. You will build a distributed platform almost from scratch and work with GPU/ML at large scale.

• 3–5+ years of experience in DevOps / MLOps / Platform Engineering. • Confident Kubernetes and/or Nomad. • Experience with large production clusters and distributed systems. • Practical experience with GPU infrastructure. • Understanding of neural network training and distributed training. • Experience with Docker / Docker Registry, CI/CD, Linux, monitoring. • Experience designing and building infrastructure from scratch. Nice to have: • Experience with HPC, supercomputers, GPU farms, Ray, and large compute clusters.

Salary: $5,000–10,000 net per month. Location: Yekaterinburg, hybrid. Employment: full-time.