Senior Machine Learning Engineer; Maternity Leave Cover

Senior Machine Learning Engineer (Maternity Leave Cover)

Department: Platform Delivery

Employment Type: Fixed Term - Full Time

Location: London

The Machine Learning Engineer role sits within Client Delivery, embedded in the Data & Analytics Consulting (DACs) team - a technical, client-facing group of Data Scientists and ML Engineers responsible for building and operationalising advanced machine learning and AI.

The Machine Learning Engineer role sits within Client Delivery, embedded in the Data & Analytics Consulting (DACs) team - a technical, client-facing group of Data Scientists and ML Engineers responsible for building and operationalising advanced machine learning and AI components across Xantura’s projects.

As an ML Engineer here, your core work is designing, training, evaluating, and product ionising machine learning models on complex, multi-source datasets from local authorities. You will engineer high-performance training pipelines, build embedding-based and sequence models, implement LLM and RAG workflows, and develop containerised model services that integrate directly into the One View platform. This includes hands-on work with model architectures, feature engineering, model optimisation, performance debugging, schema-aligned data preparation, and ML-driven interfaces.

This is a role for engineers who want to build real models, ship real systems, and solve real operational ML problems - not just prototypes. You will work directly with production data, client technical teams, and our internal engineering ecosystem to deliver AI components that are robust, scalable, and deployed into live environments.

• Design, train and optimise predictive models using advanced architectures such as gradient boosted trees, temporal models and embedding based models.
• Build robust training, evaluation and monitoring pipelines to ensure model quality, reproducibility and auditability.
• Implement feature engineering, hyperparameter tuning, model debugging and performance optimisation.
• Product ionise models so they run reliably and efficiently at scale in client environments.

• Own schema aware data flows for modelling and cohorts; validate, transform and version datasets used in training and inference.
• Manage and evolve database schemas; optimise SQL, indexing and partitioning for large training and scoring workloads.

• Lead the modelling and data engineering components of client projects alongside DACs and Business Consultants.
• Acquire and extract data from client source systems
• Build and validate cohort logic to ensure accuracy, interpretability and alignment with client needs.
• Troubleshoot and resolve complex modelling and pipeline issues throughout delivery.

• Build and integrate LLM based components including embedding pipelines, RAG workflows and text analysis models.
• Develop and deploy agentic and multicomponent AI systems using modern ML frameworks.
• Engineer high performance NLP and sequence models for information extraction, classification and risk prediction.

• Configure advanced One View components linked to modelling outputs such as risk logic, summaries and scoring pathways.
• Contribute modelling innovations, performance insights and engineering improvements back into the platform.

• Act as an SME for machine learning, AI and model engineering within DACs.
• Mentor DACs on Python, modelling best practice, data engineering fundamentals and debugging approaches.
• Produce documentation, templates and reusable components to raise engineering standards across delivery.

We'd love to hear from you if you have:

• Strong Python engineering skills and experience with modern ML frameworks
• Practical experience training and evaluating models (tree based, temporal, embedding/NLP or LLM based)
• Ability to build reproducible training and evaluation pipelines
• Experience containerising and deploying models (e.g., Docker, Fast API)

• Strong SQL and experience working with relational databases
• Understanding of schemas, data transformations and (ideally) DBT
• Experience preparing data for model training and scoring

• Working with embeddings, vector databases or RAG style workflows
• Experience applying NLP or sequence models to real world datasets