Kroo Bank is a digital first bank focused on using technology, data, and innovation to enhance banking experiences. As a Data Scientist, you will help the company leverage data across various business areas, building and deploying data science solutions to support strategic decision making and improve customer experiences.
Responsibilities:
- Build and iterate on statistical and machine learning models to solve business problems across areas such as credit risk, fraud, customer engagement, and operational efficiency
- Partner with stakeholders to define problem statements, success metrics, data requirements, and practical implementation plans
- Conduct data exploration and feature engineering to uncover drivers of outcomes and improve model performance and interpretability
- Develop robust evaluation frameworks, including appropriate baselines, validation strategies, monitoring metrics, and model performance reporting
- Support deployment of models into production in collaboration with Engineering, contributing to reproducible pipelines and model documentation
- Monitor models in production, identify performance drift, propose improvements, and support ongoing recalibration or retraining where required
- Apply probability and statistical inference to design experiments, interpret results, and provide clear recommendations to stakeholders
- Contribute to high quality data practices by identifying data quality issues, supporting cleaning and normalisation approaches, and defining standards for reliable datasets
- Write maintainable, well tested Python code using common data science libraries, and follow engineering best practices appropriate for production systems
- Use SQL and dbt to extract, transform, and validate data for analysis and modelling, ensuring traceability and reliability of outputs
- Collaborate with Risk, Compliance, and Audit stakeholders to ensure data science work is appropriately governed, documented, and aligned with regulatory expectations
- Support continuous improvement across data science methodologies, tooling, and ways of working