MLOps Engineer
Location: London, UK (Hybrid – 2 days per week in office)
Day Rate: Market rate (Inside IR35
Duration: 6 months
Role Overview
As an MLOps Engineer, you will support machine learning products from inception, working across the full data ecosystem. This includes developing application-specific data pipelines, building CI/CD pipelines that automate ML model training and deployment, publishing model results for downstream consumption, and building APIs to serve model outputs on-demand.
The role requires close collaboration with data scientists and other stakeholders to ensure ML models are production-ready, performant, secure, and compliant.
Key Responsibilities
* Design, implement, and maintain scalable ML model deployment pipelines (CI/CD for ML)
* Build infrastructure to monitor model performance, data drift, and other key metrics in production
* Develop and maintain tools for model versioning, reproducibility, and experiment tracking
* Optimize model serving infrastructure for latency, scalability, and cost
* Automate the end-to-end ML workflow, from data ingestion to model training, testing, deployment, and monitoring
* Collaborate with data scientists to ensure models are production-ready
* Implement security, compliance, and governance practices for ML systems
* Support troubleshooting and incident response for deployed ML systems
Required Skills and Experience
* Strong programming skills in Python; experience with ML libraries such as Snowpark, PySpark, or PyTorch
* Experience with containerization tools like Docker and orchestration tools like Airflow or Astronomer
* Familiarity with cloud platforms (AWS, Azure) and ML services (e.g., SageMaker, Vertex AI)
* Experience with CI/CD pipelines and automation tools such as GitHub Actions
* Understanding of monitoring and logging tools (e.g., NewRelic, Grafana)
Desirable Skills and Experience
* Prior experience deploying ML models in production environments
* Knowledge of infrastructure-as-code tools like Terraform or CloudFormation
* Familiarity with model interpretability and responsible AI practices
* Experience with feature stores and model registries