Job Description

Location: This role can be based from anywhere in England, with a minimum of once a month travel to our London office

Who are we?

We’re the original pioneers in connected commerce marketing. Since 2008, we’ve been partnering with major retailers, powering global brands, and building meaningful connections with shoppers.

We simplify the mind-boggling complexity of today’s retail media landscape. We deliver impactful campaigns that connect with people where it matters. We create seamless and personalised shopping experiences. Above all, we deliver amazing results for our partners, driven by our unshakeable desire for growth. Time after time, we change the game. SMG is home to a world-class suite of commerce advertising capabilities powered by data and cutting-edge technology. We constantly push ourselves, our tech and our industry to discover innovative new ways to connect, sell and grow.

About the role

The MLOps Engineer owns how machine learning runs in production at SMG. Working within the Data function, the role takes models developed by our data scientists and turns them into dependable, monitored, reproducible production systems behind SMG's Core Intelligence Services - the forecasting, optimisation and recommendation capabilities that power our retail media networks. This is a platform role.

You will design and own the ML platform, tooling and operational standards that the wider Data function builds on, define how deployment, monitoring and retraining are done, and set the engineering bar by example. The outputs of these systems inform commercial decisions for leading retail partners and their advertisers, so reliability, observability and trust are the core of the job. At SMG this means working with rich, high-volume commerce media data across multiple retailers, in an environment where you shape the platform rather than inherit one.

What you’ll do

ML platform & tooling

• Design, build and own the platform that takes models from development to production: packaging, versioning, model registry, CI/CD for ML, and the shared tooling that data scientists and engineers rely on to ship reliably.

Deployment & serving

• Own how models are deployed and how predictions reach their consumers - batch and online serving, rollout and rollback, inference cost, latency and reliability - working closely with DevOps Engineering.

Monitoring & observability

• Own how we know models remain healthy in production: data-quality and drift monitoring, leading indicators of degradation rather than only lagging metrics, actionable alerting, and clear operational ownership.

Retraining & reproducibility

• Establish reproducible training and a governed retraining lifecycle - evaluation against the incumbent model, promotion criteria and version control - so that model updates are routine and safe.

ML data quality

• Ensure the integrity of the data feeding models: training-data validation, feature and label integrity, leakage and train/serve skew checks, and consistent feature logic across training and inference.

Standards & technical leadership

• Define - not just follow - the ML engineering standards across the Data function: reproducibility, testing, model review and documentation. Partner day to day with data scientists and data engineers, and explain model behaviour clearly to commercial stakeholders and clients.

What you'll bring

Essential

• Extensive commercial experience in ML engineering, ML platform or MLOps roles, or in data/platform engineering with substantial production ML exposure.
• Proven experience taking models into production and keeping them there - deployment, monitoring, retraining and rollback - in commercial systems with real users and real consequences. Academic and personal projects are welcome context, but are not a substitute.
• Strong platform engineering foundations: cloud (Azure and/or AWS), containers, infrastructure as code, CI/CD, and workflow orchestration (Airflow, Databricks Workflows or similar).
• Hands-on with ML lifecycle tooling - experiment tracking, model registry (MLflow or equivalent), and CI/CD for models rather than only for applications.
• Strong software engineering fundamentals: production-level Python, testing, version control and code review. You write high-quality, secure, maintainable code others can build on.
• Deep understanding of how models fail in production - drift, train/serve skew, leakage, data quality - and how to detect and respond to each.
• Practical experience with modern data platforms, Databricks and Snowflake especially, and close collaboration with data engineering on the data that feeds models.
• Able to explain model behaviour and operational risk clearly to non-technical stakeholders and clients.
• Able to operate independently: owning delivery end to end, making sound technical calls with light direction, and raising the bar for those around you.

Desirable

• Exposure to at least one of forecasting, optimisation or recommendation systems, or clear aptitude to pick these up quickly.
• Feature stores and feature platforms, including the judgement of when they are and are not worth building.
• Experience in a lean team where you have built breadth alongside depth.
• Degree in Computer Science, Engineering, or a related discipline, or equivalent practical experience.
• Strong platform, backend or data engineers who have worked closely with production ML are very welcome to apply.

We're looking for people who enjoy the buzz of change, the satisfaction of building something better, and the joy of working with a close-knit, values-driven team. If you love variety, thrive in a fast-paced environment, and embrace change with energy, this could be your right role.

Why SMG?

At SMG, we hire for the future, which is fast-moving and changing shape. Do you have the potential to help shape our business? We’re looking for brilliant, diverse talent who want to grow with us - people who are curious, ambitious, and eager to learn, whether as specialists or across teams.

We value those who take ownership of their growth and bring fresh perspectives. That’s why we’re committed to equity, inclusion, and building a place where everyone feels empowered to grow. At SMG, it’s not just about filling a role but building the future together.

• 10% discretionary bonus
• £1,800 yearly wellbeing fund (on top of your salary!)
• Free Headspace subscription
• £500 yearly “Uni Fund” for learning
• 4 extra Wellbeing Days off per year
• Annual Summer conference + year-round celebrations
• 4pm finishes every Friday
• Flexible and hybrid working

Explore all our benefits here

Useful Links Before You Apply

• Learn about our agencies
• Our values
• Our hiring process
• EDI at SMG
• FAQs

Our full data retention policy can be found here.