Role Summary The Data Scientist is responsible for making statistically sound decisions about models and data.
This role focuses on problem formulation, model selection, evaluation, and interpretation, ensuring that models are valid, robust, and aligned with business objectives.
Key Responsibilities
Problem Framing & Modelling
- Translate business problems into statistical or machine learning problems
- Select appropriate modelling approaches and algorithms
- Design features in collaboration with Data Engineers
Model Development & Evaluation
- Train, tune, and compare models
- Define and justify evaluation metrics and validation strategies
- Perform error analysis and model diagnostics
Statistical Reasoning & Analysis
- Analyse bias, variance, overfitting, and data leakage
- Assess robustness, stability, and uncertainty of models
- Reason about correlation vs causation where applicable
Communication & Decision Support
- Interpret model results and trade-offs for technical and non-technical stakeholders
- Recommend whether a model is fit for production or not
- Clearly communicate uncertainty, assumptions, and limitations
Required Skills & Qualifications
- Solid foundation in:
- Probability and statistics
- Linear algebra
- Experience with:
- Machine learning model training and evaluation
- Experiment design and analysis
- Proficiency in Python and common ML libraries
- Ability to explain probabilistic outcomes clearly and precisely
Accountability & Success Criteria
- Models are statistically valid and meet defined performance criteria
- Evaluation methodology is sound and defensible
- Modelling decisions are explainable and aligned with business goals
Explicit Non‑Responsibilities
- Choosing model architectures or algorithms
- Hyperparameter tuning and model selection
- Statistical validation of modelling assumptions