MANAGER ARTIFICIAL INTELLIGENCE PORTFOLIO DELIVERY

Job purpose

The Manager, AI Portfolio Delivery leads Data Scientists and Feature Engineers, owning technical quality and delivery. The role directs modelling and feature engineering, reviews code and evidence, resolves complex problems and coaches specialists. It coordinates portfolio delivery and engineering handover while preserving independent validation and formal approval.
Principal responsibilities
Set the technical direction and delivery plan for Data Scientists and Feature Engineers, aligning work with business priorities, data readiness and available capacity.
Translate business problems into modelling objectives, baselines, measurable acceptance criteria and feasible technical solution designs.
Lead selection and review of statistical and machine learning methods, testing assumptions, suitability, complexity and alternatives before development proceeds.
•

Direct feature design, transformation and reusable pipelines; review data quality, lineage, leakage risks and consistency between training and production inputs.
•

Review Python and SQL code, experiments and pipeline designs; require reproducibility, version control, automated tests and peer review before handover.
•

Challenge model results using appropriate test splits, benchmarks and error analysis; assess calibration, explainability, fairness and robustness where relevant.
•

Guide retrieval, prompting and evaluation for generative AI and bounded agent workflows; test grounding, source permissions, tool limits and human oversight.
•

Resolve complex modelling and feature-pipeline problems through hands-on analysis, code review and targeted prototyping with the team.
•

Define technical handover requirements with ML Engineering and ML Ops, covering scoring consistency, interfaces, performance, monitoring and rollback readiness.
•

Review drift, model-quality and feature-quality evidence with Model Owners; recommend remediation or retraining and route material changes through required gates.
•

Ensure complete development evidence, limitations and control documentation; address validation findings without substituting team review for independent validation.
•

Manage portfolio scope, resources, budgets, dependencies and supplier technical deliverables; report quality, delivery risks and benefits against agreed baselines.
•

Coordinate business acceptance and adoption with business owners and the AI Business Analyst & Adoption role; distinguish measured value from forecasts.
•

Set staff objectives, review technical performance and coach Data Scientists and Feature Engineers; strengthen specialist capability, shared practices and succession.
•

Establish and continuously improve standards for model development, feature engineering, experimentation, coding practices, documentation, and quality assurance to promote consistency and engineering excellence across the AI portfolio.
•

Collaborate with Data Engineering, ML Engineering, MLOps, ICT, Architecture, Cybersecurity, and business stakeholders to ensure successful delivery, operational readiness, and long-term sustainability of AI solutions.
•

Monitor portfolio performance, delivery milestones, technical risks, resource utilization, and dependencies, implementing corrective actions to maintain delivery commitments and quality standards.
•