MLOps Engineer

The MLOps Platform Team at Caterpillar (via Cozen Technology Solutions) is seeking an experienced MLOps Engineer to build and maintain a self-service ML platform. The role involves designing scalable pipelines, enabling data science teams, and ensuring reliability of AI/ML models in production.

• Define scalable architectures and pipelines for building, deploying, and diagnosing production ML applications
• Enable users and teams on the ML platform; troubleshoot and debug user issues; maintain documentation and training
• Collaborate with stakeholders to build a comprehensive MLOps Platform
• Design and implement cloud solutions and MLOps pipelines on AWS
• Develop standards and examples to accelerate data science productivity; run code refactoring, containerization, deployment, versioning, and monitoring (including data & concept drift)
• Automate testing, validation, and deployment of data science models
• Provide best practices and execute POCs for automated and efficient MLOps at scale

• Bachelor's degree with 5+ years experience or Master's degree with 3+ years experience
• 5+ years experience with an object-oriented programming language (Python, Golang, Java, C/C++)
• Experience with MLOps frameworks like MLflow or Kubeflow
• Proficiency in programming (Python, R, SQL)
• Ability to design and implement cloud solutions and MLOps pipelines on AWS
• Strong understanding of DevOps principles, CI/CD, and tools (Git, GitHub, JFrog Artifactory, Azure DevOps)
• Experience with Docker and Kubernetes
• Strong communication and collaboration skills
• Ability to create User Stories and Tasks from higher-level requirements

• Ability to create model inference systems with advanced deployment methods integrating with MLFlow
• Knowledge of inference systems like Seldon or Kubeflow
• Knowledge of deploying applications in Langfuse or Kubernetes using Helm and Helmfile
• Knowledge of infrastructure orchestration using CloudFormation or Terraform
• Exposure to observability tools such as Evidently AI