What are the responsibilities and job description for the MLOps Engineer position at Evlo AI?
About The Role
The role sits at the intersection of machine learning and infrastructure — owning the CI/CD pipelines, serving infrastructure, and observability tooling that get models from notebooks into production reliably. This includes GPU resource orchestration, model registries, feature stores, and automated retraining workflows.
As ML adoption grows across the engineering organization, this role directly determines how fast teams can ship models and how confidently they can run them at scale. You will work closely with ML engineers and data scientists who expect deployment to be boring, and platform leadership who expect infrastructure costs to be defensible.
Key Responsibilities
The role sits at the intersection of machine learning and infrastructure — owning the CI/CD pipelines, serving infrastructure, and observability tooling that get models from notebooks into production reliably. This includes GPU resource orchestration, model registries, feature stores, and automated retraining workflows.
As ML adoption grows across the engineering organization, this role directly determines how fast teams can ship models and how confidently they can run them at scale. You will work closely with ML engineers and data scientists who expect deployment to be boring, and platform leadership who expect infrastructure costs to be defensible.
Key Responsibilities
- Build and maintain CI/CD pipelines for ML workflows using GitHub Actions, Argo Workflows, or Kubeflow, covering training, evaluation, and deployment stages
- Deploy and scale model serving infrastructure on Kubernetes (KServe, Seldon, or Triton Inference Server), including autoscaling, canary rollouts, and rollback strategies
- Implement model monitoring and observability using tools like Prometheus, Grafana, Evidently, or Arize — tracking latency, drift, and data quality with automated alerting
- Design and operate feature pipelines with Feast or Tecton, ensuring consistency between offline training and online serving paths
- Manage model versioning and lineage through MLflow or SageMaker Model Registry, with reproducible training runs and audit trails
- Optimize GPU utilization and cloud costs across training and inference workloads on AWS, GCP, or Azure
- Partner with data science and ML engineering teams to productionize new models, defining deployment standards and platform guardrails
- 3–6 years of experience in MLOps, platform engineering, or DevOps with a strong ML focus
- Deep hands-on experience with Kubernetes, Docker, and Terraform (or equivalent IaC) in production environments
- Proficiency in Python for tooling and pipeline development; comfort reading and debugging PyTorch or TensorFlow training code
- Experience operating ML platforms or components in production: MLflow, Kubeflow, SageMaker, Vertex AI, or Databricks
- Solid understanding of model serving patterns (batch vs. real-time inference), monitoring for drift, and retraining automation
- BS/MS in Computer Science, Engineering, or equivalent practical experience
- Bonus: Experience with distributed training (Ray, Horovod), LLM inference optimization (vLLM, TensorRT-LLM), or building internal ML platform products used by multiple teams