Nov 2025 — Present
Gas & Power Trading Analyst — North American Gas
Equinor · Stamford, CT
- Built a full-stack PJM power-price forecasting application end to end — Python backend, React front end — delivering day-ahead forecasts and analytics to the trading desk.
- Developed a lower 48 storage flow model in Python, producing weekly nowcasts of EIA reported change in storage, as well as next week forecast, and end of season final storage estimates, with a 1.1 bcf/d average error term.
- Engineered Python/Polars ETL pipelines on Azure Databricks turning weather, price, and market data into model-ready datasets feeding live applications.



