Job Responsibilities

• Design, train, and deploy time-series and tree-based models (e.g., XGBoost, CatBoost, LightGBM, sklearn) to predict fair market value and forecast volatility.
• Optimize models to improve MAE, PnL, and business outcomes in live environments.
• Build and harden the data layer: ingest and reconcile auction feeds, marketplace listings, and private-sale records.
• Handle real-world data challenges, including splits, duplicates, stale pricing, and zero-compensation scenarios.
• Maintain robust model orchestration using tools like Airflow, feature stores, and real-time inference endpoints.
• Investigate market microstructure dynamics (e.g., extended bidding, buyer premiums, cash advances).
• Translate quantitative insights into pricing logic that feeds directly into production models.
• Expose model outputs via public APIs and integrate with internal applications (e.g., price alerts, dashboards).
• Collaborate with product and data teams to deliver insights that power decision-making.

Job Requirements

• 9+ years in quantitative finance, machine learning engineering, or similar fields.
• Demonstrated expertise in forecasting and time-series modeling, ideally for illiquid or auction-based assets.
• Proven ability to move from Jupyter notebooks to production, with strong understanding of CI/CD, testing, and model monitoring.
• Hands-on proficiency with:
  • Programming: Python, SQL
  • ML Libraries: XGBoost, CatBoost, LightGBM, sklearn
  • Data Tools: Airflow, dbt, Postgres, Spark
  • Cloud: AWS (S3, ECS, Lambda)
  • DevOps: GitHub Actions, CI/CD pipelines
• Strong communication skills: capable of explaining heteroscedastic noise and complex statistical phenomena to technical and non-technical audiences alike.

Posted By