Data Scientist III

The Data Scientist - Level 3 is responsible for leading predictive modelling and experimentation to drive key business metrics around risk and impact. The role combines hands-on modelling, AI agent/ML system development, and stakeholder collaboration to turn complex data into actionable strategies. This is an IC role with mentoring juniors.

Responsibilities:

• Own end-to-end predictive modelling initiatives for business-critical metrics related to risk, impact, and growth, including problem framing, data strategy, modelling, and deployment.
• Design and execute segmentation, underwriting, and matching algorithms to improve customer targeting, pricing, and decision automation.
• Lead experimentation and post-modelling evaluation, including A/B tests, backtests, and lift analyses, ensuring robust, statistically sound results.
• Build, train, and iteratively improve ML models (supervised, unsupervised, and forecasting) using modern libraries and MLOps practices.
• Develop and train AI agents or decisioning systems that integrate with products and workflows to automate and optimise business processes.
• Translate ambiguous business or product problems into clear analytical problem statements, hypotheses, and measurable success metrics.
• Partner with product, engineering, risk, and operations to embed models into production systems and monitor performance over time.
• Communicate insights, tradeoffs, and recommendations clearly to senior stakeholders through dashboards, presentations, and concise written summaries.
• Mentor junior data scientists and analysts, promoting best practices in experimentation, coding standards, and model governance.
• Strong fundamentals in statistics and probability with demonstrated experience in mathematical and statistical modelling for real-world problems.
• Proven ability to build and design ML algorithms, including feature engineering, model selection, hyperparameter tuning, and performance optimisation.
• Hands-on implementation skills with a basic understanding of software/data engineering concepts (version control, testing, modular code, CI/CD for models).
• Expertise in Python (or similar), SQL, and modern data science tooling (e. g., Jupyter, pandas, and scikit-learn; exposure to Spark or cloud platforms is a plus).
• Experience with segmentation, matching, recommendation, or risk/underwriting models in production environments. Ability to independently research, evaluate, and apply state-of-the-art statistical/ML/AI techniques to new problem domains.
• Strong business problem translation skills: comfort working with incomplete information and iteratively refining scope with stakeholders.
• Excellent communication skills with the ability to present complex analyses simply and persuasively to technical and nontechnical audiences.

Requirements:

• 4-7 years of relevant experience in data science, machine learning, predictive modelling, or applied statistics, with a track record of owning high-impact projects end-to-end.
• Master's degree (or higher) in statistics, mathematics, computer science, engineering, or a related quantitative discipline, or a related quantitative discipline from a Tier-1 institution.
• Prior experience in a data scientist role or equivalent responsibility in a high-growth or product-led organisation.
• Strong foundations in statistics and probability, with hands-on experience applying mathematical and statistical modelling to real-world business problems.
• Proven experience designing and building machine learning models, including feature engineering, model selection, hyperparameter tuning, validation, and performance optimisation.
• Strong proficiency in Python, SQL, and modern data science tools such as Jupyter, pandas, and scikit-learn.
• Exposure to Spark, cloud platforms, or MLOps tools would be an advantage.
• Experience developing and deploying segmentation, matching, recommendation, risk, or underwriting models in production environments.
• Working knowledge of software and data engineering practices, including version control, testing, modular coding, model deployment, and CI/CD.
• Ability to independently research, evaluate, and apply emerging statistical, machine learning, and AI techniques to new problem areas.
• Strong business problem-solving skills, with the ability to work with incomplete information and translate ambiguous business requirements into measurable analytical problems.
• Excellent communication and stakeholder-management skills, with the ability to present complex analyses and recommendations clearly to technical and non-technical audiences.
• Ability to mentor junior data scientists and analysts while continuing to operate as a hands-on individual contributor.