At Green Line Commodities GmbH, I research how EU RED II/III and the German GHG quota reprice fuels and feedstocks. I turn that research into written market updates and own a Python framework that decomposes daily price observations into policy, fundamental, and residual drivers.
At Louis Dreyfus Company, I built supply-demand balances and price forecasts for a physical wheat, maize, and cotton book, and priced imports using import-parity and landed-cost models. My 2025 flood-loss forecast was 200,000 bales against a market consensus of about one million; the harvest confirmed it, earning a Merit Award for Excellence in Forecasting.
I also rebuilt daily position and risk reporting in Python and Power BI, cutting reporting time by 40%. My analysis contributed about $11 million of platform performance in 2025.
I published a live ERCOT day-ahead power-price model showing when gas sets prices and when solar and wind push them down. Its backtest error was $5.21/MWh versus $7.50 for a persistence benchmark; I also published a GB wind-curtailment model with bootstrap confidence intervals and unit tests.
