About the job
As a Credit Data Scientist at Tyme, you will leverage data analytics, feature engineering, and experimentation to enhance credit decision-making and optimize portfolio performance across our diverse lending products and markets. Your role will encompass everything from data exploration to the deployment of production-ready features, alongside continuous monitoring and impact assessment.
Key Responsibilities
· Analyze customer, bureau, transactional, and repayment data to identify key risk factors, loss drivers, approval rates, and customer outcomes.
· Develop and refine credit risk features and model inputs (such as behavioral signals and affordability proxies), collaborating closely with senior modellers and the engineering team.
· Contribute to the advancement and enhancement of predictive models using cutting-edge machine learning techniques, focusing on robustness, stability, and deployment readiness.
· Design, implement, and assess credit policy experiments (including cut-offs, limits, and pricing/risk trade-offs), followed by thorough post-implementation evaluations.
· Establish monitoring mechanisms for model/policy performance and feature integrity (addressing drift, stability, segment performance, and data quality checks).
· Assist in portfolio analytics by conducting vintage analyses, roll-rate assessments, migration tracking, early warning indicators, collections funnel analytics, and loss driver investigations.
· Collaborate with the Data/Engineering teams to enhance data definitions, quality, lineage, and reproducibility of pipelines; document feature logic and assumptions effectively.
· Contribute to the governance documentation, including model inputs, feature catalogs, monitoring evidence, and change logs.
