MLOps Engineer

Hi — I’m Gleb, an MLOps/DevOps engineer with 4+ years of experience building and scaling ML infrastructure. Recent work: • Built GPU-powered, multi-modal serving pipelines for 30k+ daily users. • Deployed Kubernetes-native GPU clusters, cutting hosting costs by 50%. • Set up Grafana Cloud dashboards and alerts, reducing on-call guesswork by over 40%. What do I do? • Bridge research and production environments. • Build reproducible development setups with Nix and dev containers. • Code in Python, Rust, Go, C++, and TypeScript. • Keep codebases clean with pre-commit hooks and conventional commits. • Package applications with Docker. • Orchestrate deployments with Kubernetes, Kustomize, and Helm. • Set up infrastructure using Terraform, Ansible, and GitOps via Argo CD. • Automate testing and deployments with CI/CD pipelines. • Train models with PyTorch and Lightning; track experiments using Weights & Biases. • Deploy those models with NVIDIA Triton Inference Server. • Manage data with Supabase and PostgreSQL. • Build asynchronous pipelines with Celery and Kafka. • Develop APIs with FastAPI and grpc-go. • Monitor metrics, logs, and traces through Grafana Cloud. What domains do I know? • Speech processing and synthesis. • Natural language processing. • Finance and quantitative modeling. How do I work? I focus on measurable wins: lower latency, higher uptime, reduced costs, and better user retention. Contact: 📧 khaykingleb@gmail.com  🌐 khaykingleb.com Skills: Python, Rust, Go, SQL, TypeScript, Kubernetes, Docker, Terraform, AWS, GCP, NVIDIA Triton, PyTorch, Lightning, Grafana