AVP/VP, Data Scientist, Data Management Office
At a glance
VP-level Data Science role at Sumitomo Mitsui Banking Corporation. Singapore.
Pay not stated Data Science salary
dataroles summary, based on the employer's posting.
What you'll do
- Judge whether AI, machine-learning and data-science models have sound aims, assumptions, inputs and logic
- Inspect how features are built and whether data problems, leakage or wrongly specified inputs could affect results
- Use performance measures and diagnostic tests to examine accuracy, stability, sensitivity, robustness and drift
- Apply SHAP, LIME and other interpretability approaches, then prepare reasoned model-validation reports
- Review LLMs, retrieval-augmented systems, agent workflows and prompt methods; contribute to exploratory AI/ML proofs of concept
What you bring
- At least four years of practical work in model development, validation, quantitative analytics or AI/ML evaluation in finance or a regulated setting
- Strong grounding in machine learning, AI, statistical modelling and validation methods, including feature design and data checks
- You can assess model performance and diagnostics, spot design weaknesses, and write high-quality validation reports
- You understand model interpretation methods and Responsible AI concerns, including fairness, transparency and robustness
Who this fits
You’ll suit work that calls for independent technical judgment and careful written analysis. You’ll coordinate with modelling specialists across business and functional groups. A postgraduate degree in a quantitative discipline is preferred.
From the employer
Responsibilities
• Independently assess AI/ML/data science model purpose, assumptions, features, data inputs, and logical soundness.
• Evaluate feature engineering, data quality, and detect issues such as leakage or mis-specified inputs.
• Evaluate model performance using suitable metrics, diagnostic tests, and validation methodologies.
• Assess stability, robustness, sensitivity analysis, susceptibility to adversarial attacks and model or concept drift.
• Apply model explainability methods such as SHAP, LIME and other interpretability techniques.
• Produce comprehensive, well-reasoned Model Validation Reports.
• Evaluate AI/ML models, LLMs, retrieval-augmented systems, agentic workflows, and prompt-engineering methods.
• Ensure validation standards align with Responsible AI principles including fairness, transparency, and robustness.
• Collaborate with data scientists and model developers across business and functional teams to understand modelling intent, design rationale, and underlying assumptions.
• Contribute to exploratory AI/ML proof‑of‑concept (POC) initiatives to deepen technical understanding, enhance validation methods, and support innovation within DMO.
Requirements
• Preferably a postgraduate degree in Data Science, Statistics, Mathematics, Analytics, Computer Science, or quantitative discipline.
• At least 4 years of hands‑on experience in model development, model validation, quantitative analytics, or AI/ML evaluation within financial institutions or similarly regulated environments.
• Strong theoretical and practical knowledge of machine learning, AI, statistical models, and model validation techniques.
• Strong understanding of feature engineering, feature selection, and data quality checks.
• Proficiency in evaluating model performance and diagnostics across statistical, ML, and AI models.
• Understanding of explainability techniques, including SHAP, LIME, and other model interpretation methods.
• Analytical skills to identify modelling weaknesses, design flaws, and performance gaps.
• Strong reporting skills to produce high-quality validation deliverables.
• Familiarity with Responsible AI concepts such as fairness, transparency, and robustness.
Data Science work from LinkedIn, X, Reddit and Upwork, in one list.
Each morning you get everything that fits, strongest first.
Data Science
Find me workBy the team behind dataroles.