MLOps Engineer

The MLOps Engineer will build and maintain the infrastructure, automation, and deployment pipelines required to run machine learning models reliably in production.

The role combines software engineering, cloud infrastructure, CI/CD, model lifecycle management, monitoring, and operational reliability.

The successful candidate will work closely with ML engineers, data scientists, and platform engineers to make model training, testing, deployment, monitoring, and retraining reproducible and scalable.

Health Insurance

Private healthcare coverage supporting physical and mental wellbeing.

Flexible Work

Hybrid working arrangements with flexibility around work location and schedules.

Paid Time Off

Paid annual leave to support rest and work-life balance.

Learning & Budget

Dedicated budget for AI/ML courses, technical certifications, conferences, and professional development.

• Retirement Plan
• Wellness / Stipend

Wellbeing allowance for approved health, fitness, and wellness activities.

• 02 Responsibilities
• 01 Build and maintain automated ML training and deployment pipelines.
• 02 Implement CI/CD workflows for machine learning applications.
• 03 Manage model versioning, artefact storage, and experiment tracking.
• 04 Deploy models using cloud infrastructure and containerisation.
• 05 Use Docker and Kubernetes or equivalent orchestration technologies.
• 06 Monitor model performance, availability, latency, and resource usage.
• 07 Implement alerts for model drift, pipeline failures, and service degradation.
• 08 Automate model testing, validation, release, and rollback procedures.
• 09 Manage infrastructure configuration and deployment environments.
• 10 Apply access controls, secrets management, and security best practices.
• 11 Troubleshoot production incidents and improve system reliability.
• 12 Document architecture, deployment procedures, and operational processes.

What we evaluate against.

• ML Training and Deployment Pipeline Automation
• Must have
• Build automated ML training and deployment pipelines
• Implement CI/CD workflows specifically tailored for machine learning applications
• Automate model testing, validation, release, and rollback procedures
• Nice to have
• Experience with CI/CD tools (e. g. Jenkins, Git Lab CI, Git Hub Actions, Argo) applied to ML workflows
• Track record of reducing manual steps in model release processes
• Containerisation and Cloud Infrastructure Deployment
• Must have
• Deploy models using cloud infrastructure and containerisation
• Use Docker and Kubernetes or equivalent orchestration technologies
• Manage infrastructure configuration and deployment environments
• Nice to have
• Hands-on experience with a major cloud provider (AWS, GCP, or Azure)
• Experience with infrastructure-as-code tools for environment provisioning
• Model Versioning and Experiment Lifecycle Management
• Must have
• Manage model versioning and artefact storage
• Implement and maintain experiment tracking systems
• Ensure training, testing, and deployment processes are reproducible and scalable
• Nice to have
• Familiarity with tools such as MLflow, DVC, or similar model/experiment registries
• Experience coordinating model lifecycle work with ML engineers and data scientists
• Production Model Monitoring and Alerting
• Must have
• Monitor model performance, availability, latency, and resource usage
• Implement alerts for model drift, pipeline failures, and service degradation
• Nice to have
• Experience with observability stacks (e. g. Prometheus, Grafana, ELK) applied to ML systems
• Understanding of statistical approaches to detecting model and data drift
• Operational Reliability, Security, and Incident Response
• Must have
• Troubleshoot production incidents and improve system reliability
• Apply access controls, secrets management, and security best practices
• Document architecture, deployment procedures, and operational processes
• Nice to have
• Experience with incident response, postmortems, or on-call rotations
• Familiarity with secrets management tools (e. g. Vault, cloud KMS) and least-privilege access design
• 04 Team dynamics Team Structure

Cross-functional collaboration with data scientists, AI engineers, software developers, and product teams.

Operating Rhythm

Weekly sprint planning, model performance reviews, technical discussions, and progress updates.

Data-driven collaboration focused on experimentation, knowledge sharing, and joint problem-solving.

Communication

Clear, evidence-based communication of model performance, technical risks, and recommendations.

Decision Making

Decisions supported by model evaluation results, experimental evidence, data quality, and business requirements.

Documentation Culture

Documented experiments, model versions, evaluation metrics, technical decisions, and implementation processes.

• 05 Company culture
• Ownership

Employees take responsibility for AI initiatives, technical decisions, and measurable business outcomes.

Feedback

Teams share constructive feedback through model reviews, code reviews, and product retrospectives.

Transparency

AI decisions, model limitations, experiment results, and project risks are communicated openly.

Teams explore emerging AI/ML technologies to build useful, scalable, and differentiated products.

Learning Culture

Employees continuously develop skills in machine learning, data science, model evaluation, and responsible AI.

Data scientists, ML engineers, software engineers, and product teams work together toward shared goals.

• 06 What success looks like
• First 30 days

The engineer becomes familiar with existing ML pipelines, infrastructure, deployment environments, and current monitoring setups, identifying gaps in automation, reliability, or documentation.

They begin contributing small fixes or improvements to CI/CD workflows and collaborate with ML engineers and data scientists to understand model handoff pain points.

90 days

The engineer has automated or improved at least one significant part of the training-to-deployment pipeline, strengthened model versioning and experiment tracking practices, and implemented or refined monitoring and alerting for drift and service health.

They are actively participating in incident response and have documented key operational procedures.

Outcomes

Models move from development to production reliably, reproducibly, and with minimal manual intervention, supported by robust CI/CD, containerised deployments, and strong observability.

Production incidents are resolved quickly, security and access practices are consistently applied, and the ML platform scales smoothly to support growing model workloads.

#J-18808-Ljbffr