MLOps Engineer — Evlo AI, New York, NY

About The Role

The role owns the infrastructure that makes ML possible at scale: training pipelines, model registries, CI/CD for ML, and the serving stack that keeps models fast, reliable, and observable in production.

The team sits at the intersection of platform engineering and machine learning — working directly with data scientists and backend engineers to turn experimental models into dependable, high-throughput systems.

Key Responsibilities

• Build and maintain end-to-end ML pipelines (training, validation, deployment) using Kubeflow, Airflow, or Dagster

• Own the model serving infrastructure — containerized inference services on Kubernetes with GPU scheduling, autoscaling, and sub-100ms latency targets

• Design CI/CD workflows for ML: automated testing, model versioning with MLflow, canary deployments, and one-click rollback

• Implement monitoring and observability for production models: data drift detection, latency/throughput dashboards, and alerting with Prometheus and Grafana

• Manage feature stores and data pipelines to guarantee training/serving consistency across environments

• Optimize inference cost and performance — quantization, batching strategies, GPU utilization, and model compilation (TensorRT, ONNX)

• Establish MLOps standards and tooling; run postmortems and drive reliability improvements across the ML platform

What We Are Looking For

• 3–6 years of experience in MLOps, ML infrastructure, or backend/platform engineering supporting ML systems in production

• Strong Python and/or Go skills; solid software engineering fundamentals including testing and code review discipline

• Hands-on production experience with Kubernetes, Docker, and at least one major cloud (AWS, GCP, or Azure)

• Experience with ML workflow tooling: MLflow, Kubeflow, Airflow, or Vertex AI Pipelines

• Track record of operating ML systems at scale — monitoring, incident response, and performance tuning under real traffic

• BS in Computer Science, Engineering, or equivalent practical experience

• Bonus: GPU orchestration experience (NVIDIA Triton, Kubernetes device plugins), feature store experience (Feast, Tecton), or contributions to open-source MLOps tooling