You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A real-data market intelligence web app for the Great Britain power market, in the style of a commodity-analytics terminal. Single-page, static, no build step, driven entirely by public data fetched through a re-runnable ETL.
Predicted Spanish day-ahead energy demand and price with 97.5% accuracy using a range of ML and statistical time series forecasting models including XGBoost, Transformers, TFTs and SARIMA.
Self-hosted, bitemporal (point-in-time-correct) data platform for power markets: EIA-930 (all US balancing authorities) + ERCOT real-time/day-ahead prices and load into a versioned ArcticDB store, Dagster-orchestrated, served point-in-time via a read API.
Decision engine for flexible electricity demand: finds cheap power windows, validates whether they can be trusted, and audits decisions through replay.
Pricing energy options (focus on German power) with MC and jump-diffusion mean-reversion model, including stochastic volatility, seasonality and regime filtering. Model parameters are calibrated on historical ENTSO-e data.
Scenario-based analysis of global oil market supply–demand balances, OPEC+ policy sensitivity and geopolitical risk drivers, using OPEC, EIA and OEUK data.
Regime-conditional volatility forecasting framework using HAR-RV as a baseline and XGBoost on either residual vol or directly on log(RV), implemented for Germany and France electricity markets. Metric: Spearman ranking. Model validation and market-neutral cross-country trading strategy.