Machine Learning Operations Engineer

Job Purpose

Develop, productionize and operate machine learning solutions across the full lifecycle, from problem framing and model development through deployment, monitoring and continuous improvement. The role bridges Data Science and MLOps, with greater emphasis on hands-on machine learning, Python engineering and practical model delivery than on deep infrastructure or platform administration.

Ideally, they are looking for candidates with a strong Data Science / Machine Learning background, with hands-on experience in developing ML models and deploying them into production.

Key Responsibilities:

• Design, develop and validate machine learning models for banking and customer-focused use cases.
• Work with business and product stakeholders to translate requirements into measurable ML solutions and success criteria.
• Prepare data, engineer features, select appropriate algorithms and conduct robust model evaluation and explainability analysis.
• Convert notebooks and prototypes into modular, tested and maintainable Python code suitable for production.
• Build reproducible training and inference pipelines for batch and, where required, real-time use cases.
• Deploy models on AWS using appropriate managed services, containers, APIs or scheduled processing patterns.
• Implement model versioning, experiment tracking, approval gates and release processes in collaboration with MLOps and platform teams.
• Monitor model performance, data quality, drift, service health and business outcomes; support retraining and remediation when required.
• Create clear technical documentation covering assumptions, methodology, validation evidence, deployment design and operational controls.
• Follow information security, data privacy, Responsible AI, model risk and audit requirements applicable to a regulated banking environment.
• Collaborate with Data Scientists, Data Engineers, MLOps Engineers, Architecture, Information Security and Model Validation teams

The ideal profile should be somewhere between a Data Scientist and an MLOps Engineer, with the following skills:

• Strong ML and Data Science background, including hands-on model development.
• Strong Python skills and experience deploying ML models into production.
• Good AWS knowledge, preferably with SageMaker experience.
• Dataiku experience is a plus, but not mandatory.
• Banking experience is preferred, but not mandatory.
• Deep expertise in Kubernetes, infrastructure engineering or platform administration is no longer a priority.

Required Skills:

• Strong hands-on Python for machine learning and production-quality software development.
• Solid knowledge of supervised and unsupervised learning, feature engineering, model selection, cross-validation, and performance metrics.
• Practical experience with common ML libraries such as scikit-learn, XGBoost, LightGBM, pandas and NumPy, deep learning frameworks where relevant.
• Good SQL skills and experience working with large, complex datasets.
• Experience productionizing models, including packaging, APIs, batch inference, containers, and automated pipelines.
• Hands-on AWS knowledge, preferably including SageMaker, S3, ECR, ECS/EKS or Lambda, Step Functions and CloudWatch.
• Experience with Git-based development, code reviews, automated testing, and CI/CD fundamentals.
• Understanding of model monitoring, data/model drift, explainability, reproducibility and model lifecycle governance.
• Ability to communicate technical findings clearly to both technical and business stakeholders.
• 6-7+ years across Data Science, Machine Learning Engineering, or applied ML roles, with evidence of models deployed into production.
• Experience delivering end-to-end ML use cases rather than working only on infrastructure or platform administration.
• Experience in banking, financial services or another regulated industry is preferred.
• Understanding of model governance, validation and documentation expectations in controlled environments is advantageous.

Key Deliverables:

• Production-ready ML models aligned with agreed business and technical requirements.
• Reproducible training, validation and inference pipelines with appropriate testing and version control.
• Deployed batch or real-time inference solutions integrated with the target AWS and enterprise environment.
• Model monitoring, alerting and retraining approach covering performance, drift and operational health.
• Complete technical, deployment, and model governance documentation to support validation, approval and production support.
• Reusable ML engineering components and practical knowledge transfer to internal teams.

Salt is acting as an Employment Business in relation to this vacancy.