MLOps Engineer (Python) - 512

MANAGERS NOTES

Engineering vs. Analysis: The role requires "engineers or developers," not Data Scientists or Data Engineers.

- Resumes should emphasize Python or GoLang for systems development rather than just data analysis or model training.

- Infrastructure Mastery: Look for deep knowledge in EKS (Kubernetes) components and AWS services like EFS, SQS, EC2, and Lambda. Candidates must demonstrate they can "create/modify" these technologies, not just utilize them.

- DevOps Crossover: Successful candidates bridge the gap between development and operations. Resumes should showcase Infrastructure as Code (IaC) tools like Terraform or CloudFormation alongside containerization with Docker.

- Operational Focus: The primary task is deploying and packaging models as services. Highlight experience in CI/CD pipelines, model orchestration, and serving models at scale using specialized compute like NVIDIA GPUs on EC2.

- Model Ownership: My team does not "own the models." Resumes that focus heavily on model architecture or feature engineering may be less relevant than those focusing on deployment, hosting, and orchestration.

Position’s Contributions to Work Group:

- The MLOps Platform Team works within the Enterprise Data and Analytics Organization at Client.

- Driving the ability to work with Internal Teams to be able to support the full life-cycle of AI and machine learning development through to beyond production.

- Helping build a platform that enables data driven decisions across the enterprise, helping teams build high-value data and AI/ML products, and enable the operationalization and reliability of all models.

- We are searching for a driven and highly skilled MLOps Engineer to join our MLOps Platform team at ServiceNow.

- The role will build the MLOps Platform, build self-service ML Development tooling, and building platform adoption.

You have ideas on how to create a great user experience for those building,- deploying, and operationalizing production quality Machine Learning models.

Typical task breakdown:

· Define scalable and secure architectures, frameworks and pipelines for building, deploying and diagnosing production ML applications

· Enable users & teams on the ML platform; troubleshoot and debug user issues; maintain user-friendly documentation and training.

· Collaborate with internal stakeholders to build a comprehensive MLOps Platform

· Design and implement cloud solutions and build MLOps pipelines on cloud solutions (e.g., AWS)

· Develop standards and examples to accelerate the productivity of data science teams.

· Run code refactoring and optimization, containerization, deployment, versioning, and monitoring of its quality, including data & concept drift

· Create way to automate the testing, validation, and deployment of data science models

· Provide best practices and execute POC for automated and efficient MLOps at scale

Interaction with team:

- Working with core team, maybe work with additional teams when needed.

- Internal only position

- Working with engineers and scrum team.

Work environment:

Onsite 2-3 days a week/ no exceptions.

Education & Experience Required:

- Bachelors degree with 5 years experience

- Master’s degree with 3 years experience

Required Technical Skills

(Required)

· 5 years of experience working with an object-oriented programming language (Python, Golang, Java, C/C etc.)

· Experience with MLOps frameworks like MLflow, Kubeflow, etc.

· Proficiency in programming (Python, R, SQL)

· Ability to design and implement cloud solutions and build MLOps pipelines on cloud solutions (e.g., AWS)

· Strong understanding of DevOps principles and practices, CI/CD, etc. and tools (Git, GitHub, jFrog Artifactory, Azure DevOps, etc.)

· Experience with containerization technologies like Docker and Kubernetes

· Strong communication and collaboration skills

· Ability to help work with a team to create User Stories and Tasks out of higher-level requirements.

Nice to Have:

· Ability to create model inference systems with advanced deployment methods that integrate with other MLOps components like MLFlow.

· Knowledge of inference systems like Seldon, Kubeflow, etc.

· Knowledge of deploying applications and systems in Langfuse or Kubernetes using Helm and Helmfile.

· Knowledge of infrastructure orchestration using ClodFormation or Terraform

· Exposure to observability tools (such as Evidently AI)

Soft Skills

(Required)

- Someone who takes the initiative on their own

- Someone who does not need to be micromanaged.