ML Ops Engineer

Global; Greater London; On-site; GBP 90,000 - 130,000; Full time

Job summary

Global in London is seeking a MLOps Engineer to build the infrastructure that brings AI/ML models into production, owning pipelines, platforms and processes for deployment, monitoring and governance at scale.

You’ll collaborate with Data Science and Product teams, mentor junior engineers and implement ML-specific CI/CD, feature stores and model registries to keep systems reliable and auditable.

Qualifications

• Operacionalised ML models into production systems.
• Experience building ML pipelines and deployment workflows.
• Strong coding in Python for production use.
• Experience with cloud ML services (AWS).
• Experience with CI/CD for ML projects.

Responsibilities

• Build automated pipelines for model training, validation and deployment.
• Implement monitoring, alerting and automated recovery for ML workloads.
• Establish controls for model lineage and audit trails; enable CI/CD for ML.

Skills

MLOps experience

Python

AWS SageMaker

Feature stores

Terraform

Docker

Cross-disciplinary comms

Tools

SageMaker

Lambda

ECS/EKS

Step Functions

Job description

Accepting Applications Until

16 October 2026

Job Description

Your New Role

MLOps Engineer

Global:IQ is the team building our new intelligence platform, turning first-party and partner data into smarter, data-led media plans across Global’s audio and Outdoor inventory.

As a MLOps Engineer at Global, you’ll build the operational infrastructure that brings AI and ML models into production. You’ll own the platforms, pipelines and processes that let our Data Science teams deploy, monitor, retrain and govern models reliably at scale—from the ground up.

Key Responsibilities

• ML Infrastructure & Deployment (40%): Build automated pipelines for model training, validation and deployment, plus model registries, feature stores and inference services, with self-serve tooling for Data Science teams.
• Model Monitoring & Operations (30%): Implement monitoring, alerting and automated recovery for ML workloads - covering latency, data quality and drift and own rollback, rollout and incident response.
• MLOps Governance & Best Practice (20%): Establish controls for model lineage, reproducibility and audit trails, and introduce ML-specific CI/CD, testing and release automation.
• Collaboration & Enablement (10%): Partner with Data Science, Data Engineering and Product, and mentor junior engineers to raise operational standards.

What You Will Love About This Role

• Think Big: This is a true AI-driven product - ML isn’t a feature, it’s the product, and your infrastructure directly enables business value.
• Own It: You’re not maintaining legacy systems - you’re establishing the MLOps patterns and standards that will scale for years.
• Keep it Simple: You’ll build pragmatic, reusable patterns that keep ML systems reliable and maintainable without over-engineering.
• Better Together: Global:IQ is a tight collaboration between technical and commercial teams.

What Success Looks Like

In Your First Few Months, You’ll Have

• Defined a clear operating model between MLOps and the teams developing models.
• Delivered an end-to-end MLOps path for at least one production use case, from model handoff through deployment, monitoring and rollout.
• Established baseline standards for model versioning, environment management and deployment.
• Implemented monitoring and alerting across operational health, data quality and model performance.

What You’ll Need

• MLOps experience: You’ve operationalised ML models in production, owning deployment, monitoring and lifecycle management.
• Strong programming: Production-quality, testable Python.
• Cloud expertise: Deep AWS knowledge (SageMaker, Lambda, ECS/EKS, Step Functions); Snowflake a plus.
• MLOps tooling: Experiment tracking and registries, workflow orchestration, model serving and feature stores.
• CI/CD & IaC: ML-specific CI/CD, Terraform, Docker and test automation.
• Cross-disciplinary communication: You translate between Data Science and Engineering and explain trade-offs to any audience.