Data & Analytics Jobs in Singapore, Singapore (Now Hiring) — 1,159 open — Page 12

Brief Job Description:

Builds and maintains the infrastructure and tooling that keeps machine learning systems reliable in production — from designing CI/CD pipelines and model deployment workflows to monitoring performance and managing model lifecycle at scale. The role works closely with ML, backend, and platform teams, contributes to automation frameworks and observability standards, and helps ensure AI models move seamlessly from experimentation to production. Requires 5+ years of MLOps or DevOps engineering experience, with a track record of operating robust ML infrastructure in production-grade environments.

Mission:

Make the path from model experiment to reliable production as fast, automated, and observable as possible.

Education:

• Bachelor’s degree or higher in Computer Science, Engineering, or a related field — or equivalent practical experience.

Duties and Responsibilities:

• Design and maintain CI/CD pipelines for ML models and services.
• Build and operate model deployment, serving, and rollback workflows.
• Implement monitoring, observability, and alerting for models in production.
• Manage the model lifecycle with MLflow: experiment tracking, versioning, registries, and reproducibility.
• Automate infrastructure and partner with ML and platform teams on standards.

Qualifications & Experience:

• 5+ years of MLOps or DevOps engineering experience.
• Strong with containers and orchestration (Docker, Kubernetes).
• Experience with CI/CD (GitLab CI) and infrastructure-as-code (Terraform or similar).
• Hands-on with a major cloud platform (Azure, AWS, or GCP).
• Hands-on experience with MLflow for experiment tracking and model registry.
• Familiarity with ML frameworks (PyTorch), workflow orchestration (Kubeflow or similar), and monitoring stacks.

Nice to Have (Not Obligatory):

• Experience serving LLMs or large models in production.
• Knowledge of feature stores and data pipeline tooling.
• Cost and latency optimisation for model serving.