MLOps Engineer

QUALIFICATIONS

• Experience deploying AI and ML systems in production. Experience working on project-based or consulting teams that deliver into client environment is preferred.
• Hands-on Kubernetes experience in production: Helm, ingress, persistent storage, autoscaling, and GPU scheduling.
• Working depth in at least one major cloud (AWS, Azure, GCP): provisioning, IAM, networking, autoscaling, and its managed AI/ML services.
• Understanding of how LLM systems behave in production: model serving and quantization, GPU memory sizing, vector databases, RAG components, and gateway and guardrail layers.
• Regular use of AI coding agents in your own production work, and experience shaping how they behave: writing tool and MCP server definitions, maintaining repository context files, building eval suites, and setting guardrails.
• Infrastructure as code and CI/CD: Terraform or Pulumi, Ansible, GitHub Actions or GitLab CI or Azure DevOps, container builds, trunk-based development, and test-driven development.
• Experience with MLOps tooling and platforms: MLflow, Kubeflow, model registries, feature stores, and at least one of Databricks, SageMaker, Azure ML, Vertex AI, Domino, or Dataiku.
• Observability for ML and LLM workloads: Prometheus and Grafana, OpenTelemetry, LLM tracing tools (LangFuse, LangSmith, Arize, or similar), drift monitoring, and cost tracking.
• Experience with common data science languages; Python, SQL, and shell scripting.
• Knowledge of the ML lifecycle (data wrangling, model selection, training, validation, deployment, retraining) and experience working day to day with data scientists.
• Familiarity with cloud data platforms such as Snowflake, Databricks, or Microsoft Fabric.
• Clear written and spoken communication with teammates, client engineers, and executives. Comfortable presenting architecture decisions and tradeoffs, and experience mentoring other engineers.

Preferred

* On-prem or hybrid infrastructure experience: GPU