Responsibilities
- Research, develop, and ship sophisticated pricing models (e.g. Tree-based, Gradient-based, Deep Learning, ensemble methods) to improve valuation accuracy for high-value assets.
- Identify new data signals — emanating from data such as auction results to social sentiment — and validate their predictive power through rigorous backtesting and experimentation.
- Partner with the Product Engineering team to refine our Risk & Underwriting models, balancing the goal of maximizing cash advances with the need to maintain a balanced lending portfolio.
- Collaborate with the Expert Pricing team and domain experts to encode market nuances into automated features and inferences, ensuring our models respect the qualitative factors collectors care about.
- Define, monitor, and lift model performance metrics and communicate the "why" behind price coverage, movements, and freshness toward becoming the ultimate liquidity platform.
Requirements
- 5+ years of experience in Data Science, with a focus on feature engineering and algorithm/model-building in the pricing, financial modeling, or marketplace dynamics domains.
- Expert-level proficiency in SQL, Python, and populate Python libraries (Pandas, Scikit-Learn, XGBoost/LightGBM, PyTorch/Tensforflow).
- Statistical rigor and deep understanding of experimental design, backtesting methodologies, and dealing with "noisy" or sparse data environments.
- Collaborative experience working in a sprint-based environment alongside Data and ML Engineers; you understand enough about model hand-offs and quality code to ensure your models are efficient and readily "production-capable".
- Proficient in developing in cloud environments (AWS, GCP) and familiarity with ML Ops infrastructure (e.g. MLFlow, workflow and container orchestration, deployment and telemetry patterns).
Nice to Have
- Advanced degree (Master’s or PhD) in a quantitative field such as Statistics, Economics, Mathematics, or Computer Science is preferred.
