Advanced Analytics: Machine Learning Engineer

Full description

Position Type Permanent Role Family Information Technology Cluster Momentum Corporate Remote Opportunity Some of the time Location - Town / City Cape Town or Centurion Introduction Through our client-facing brands Metropolitan and Momentum, with Multiply (wellness and rewards programme), and our other specialist brands, including Guardrisk and Eris Property Group, the group enables business and people from all walks of life to achieve their financial goals and life aspirations. We help people grow their savings, protect what matters to them and invest for the future. We help companies and organisations care for and reward their employees and members. Through our own network of advisers or via independent brokers and utilising new platforms Momentum Group provides practical financial solutions for people, communities and businesses. Visit us at www.momentumgroupltd.co.za Disclaimer As an applicant, please verify the legitimacy of this job advert on our company career page. Role Purpose Drive the productionisation, operationalisation, and lifecycle management of advanced analytics and machine learning models within Momentum Corporate, with the appropriate incorporation of AI-based models. The role transforms actuarial and data science prototypes into robust, scalable, auditable production systems in a regulated insurance context - partnering closely with Actuarial, Data Science, and IT to ensure reliability, explainability, and value at scale. Requirements Experience and Qualifications - Degree in Computer Science, Data Science, Engineering, Mathematics, Statistics, or related field (or equivalent experience). - 3+ years’ experience in ML engineering, production data science, or platform engineering, including hands-on deployment and operation of ML models in production. - 2+ years’ experience in regulated environments (preferably insurance/financial services): audit trails, explainability, approvals, data protection. - Demonstrable experience with model CI/CD, packaging (containers/conda), model & data versioning, and monitoring/alerting. - Experience building batch scoring pipelines (e.g., orchestrated jobs) and, where applicable, real-time inference (APIs, message queues). - Experience operating production ML workloads with appropriate monitoring, logging, alerting, and runtime optimisation. - Proficiency in Python and SQL.ML libraries: scikit-learn; familiarity with PyTorch and/or Keras (nice to have). - Cloud ML platforms: Azure ML, AWS SageMaker or Databricks; model registry/experiment tracking (e.g., MLflow). - Orchestration: Airflow / Azure Data Factory / similar. - CI/CD: Git, GitHub/GitLab/Azure DevOps pipelines. - Excellent technical and communication skills. - Knowledge of the financial products applicable, as well as the insurance and retirement fund industry would be advantageous. Duties & Responsibilities INTERNAL PROCESS - Actively participates in the design, build and deployment of the solution to a set problem statement by applying advanced analytics, with a particular focus on: - Driving the transformation of actuarial and data science prototypes into production-grade services (batch jobs and/or APIs), including packaging, testing, documentation, and release. - For mature projects and data pipelines establish Continuous Integration or Continuous Delivery (CI/CD) for ML including scalable, automated scoring pipelines and selecting fit-for-purpose architectures (scheduled batch vs. real-time API). - For new/emerging projects and data pipelines establish tactical data pipelines where appropriate. - Design, deploy, and operate ML pipelines that can leverage Microsoft Fabric artefacts developed by actuarial and other business facing specialists (Lakehouse, Notebooks, Data Pipelines, Warehouse, OneLake). - Partner with Data Engineering and IT to ensure reliable, well-documented data inputs for training, scoring, and monitoring. - Collaborate with specialists in IT to align on consistent standards and hybrid solutions. - Support the use of machine learning techniques or tools to select features, create and optimise models. Where relevant, aid in the preprocessing of structured and unstructured data, including AI-based techniques. - Prepare pipelines and deployment patterns that can integrate or leverage off of AI/LLM components (retrieval, orchestration, evaluation) where relevant. - Standardise repository structures, coding conventions, and templates for reusable model components and jobs. - Strategic implementation of ML and AI-based solutions to enhance model efficiency. - Promote reusability, maintainability, reliability, and scalability in design and development of model solutions. - Embed governance by design: model cards, validation evidence, performance baselines, approvals and sign-offs for regulated use. - Ensure explainability and auditability of production solutions (feature attributions/diagnostics appropriate to the model class and use case). - Act as the technical interface to IT for environments, network/security patterns, deployment pathways, and operational readiness. - Collaborate closely with actuaries and data scientists to production-align model design (latency, throughput, cost, explainability, stability). - Provide technical enablement to the analytics team (templates, how-to guides, code reviews) to uplift production-ready practices. - Design for cost efficiency across compute, storage, and orchestration; monitor runtime costs and optimise resource utilisation. - Support model performance by catering for monitoring data drift and re-training systems. - Recommend and implement MLOps tooling (experiment tracking, model registry, orchestration, monitoring) and evolve standards across the team. - Keep current with platform capabilities (e.g. Azure ML vs AWS vs Databricks, containerisation, orchestration services) and introduce pragmatic improvements that reduce risk and lead time. - Collaborating with cross-functional teams: build and maintain relationships with clients and stakeholders that promote cross-delivery practice solutions. CLIENT - Maintain strong relationships with Actuarial, Data Science, IT, and business owners to deliver reliable production ML services that meet agreed SLAs/SLOs. - Represent Advanced Analytics in architecture, security, and governance forums to ensure compliant deployment patterns. - Continuously improve turnaround times, stability, and observability of ML services; proactively communicate incidents and remediation. - Champion solution explainability and auditability to support business adoption and regulatory expectations. PEOPLE - Establish and socialise engineering standards for model packaging, testing, observability, and documentation. - Operate effectively in a lean, business-embedded advanced analytics team, balancing tactical delivery with the progressive maturation of engineering practices. - Coach actuaries and data scientists on production-ready coding practices (data contracts, dependency management, testing, reproducibility). - Contribute to a culture of continuous improvement and incident learning (post-mortems, blameless RCA, playbooks). - Participate in a positive work climate and culture to energise employees and give meaning to work. - Promote a culture that guides and directs best practice, fostering an environment of continuous learning, improvement and cohesiveness. - Participate in a learning and growth culture whereby information regarding successes, issues, trends and ideas are actively shared within Momentum Corporate but also at Group level, through participation in Group-wide Forums - Participate in performance management frameworks in line with Momentum Corporate’s guidelines. - Encourage innovation, change agility and collaboration within the team. FINANCE - Design for cost efficiency across compute, storage, and orchestration; monitor and optimise runtime costs of pipelines and model endpoints. - Implement controls that reduce operational and compliance risk (access, secrets, data minimisation, retention). - Track and report service reliability metrics (uptime, latency, failure rates) and cost KPIs relevant to production ML operations. - Optimally contribute to the budget design for the area, including the motivation for expenditures and implementation of financial regulations. - Follow risk management, governance, and compliance policies in own practice area, to identify and manage risk exposure liability. Competencies - End-to-End Ownership & Delivery: Takes accountability for production ML services from deployment through operation, monitoring, and continuous improvement. - Systems Thinking & Engineering Rigor: Designs solutions with reliability, observability, scalability, and full lifecycle management in mind. - Pragmatic Problem-solving: Favors simple, robust, and maintainable solutions that balance technical excellence, risk, maturity of business area and business value. - Collaboration & Technical Influence: Effectively bridges Actuarial, Data Science, IT, and business stakeholders through clear, credible technical communication. - Governance & Compliance Mindset: Embeds auditability, explainability, security, and regulatory compliance by design in all production solutions. - Continuous Improvement & Innovation: Drives automation, learning from incidents (post-mortems), and incremental innovation to improve reliability and speed to value. - Business & Stakeholder Acumen: Demonstrates understanding of the group insurance environment and a strong commitment to stakeholder outcomes and service excellence. - Professional Impact & Team Contribution: Shows self-awareness, inclusiveness, and a collaborative mindset; contributes to growing team capability and supporting change.