Forecasting • Causal Inference • Python / R / SQL / Stata
I work at the intersection of forecasting, causal inference, and applied economics. My focus is on building transparent, end‑to‑end analytical workflows: data construction, model design, identification strategy, and interpretation. The goal is simple—produce results that are reproducible, defensible, and useful.
This site collects that work: forecasting pipelines, causal evaluations, and the underlying Python, R, and SQL that make the analysis possible.
I build forecasting models and causal inference frameworks grounded in econometric rigor. My work includes time‑series forecasting, difference‑in‑differences, event studies, and synthetic control designs applied to real economic and policy questions. Each project includes full data preparation, model construction, and methodological documentation.
The approach is consistent across domains: clear assumptions, transparent uncertainty, and models designed to generalize. Whether the problem is a commodity forecast, a demand model, or a policy evaluation, the objective is the same—extract signal from data and quantify how the world is likely to change.
Silver price forecast, April 2026 — a Monte Carlo forecasting model incorporating macro factors, cross‑asset relationships, and recent volatility structure. The model runs 10,000 simulated paths and calibrates drift and variance using a rolling six‑month window.
2026 FIDE Candidates Tournament — a performance model combining Elo trends, recent form, head‑to‑head records, and historical tournament outcomes. The model produces player‑level expected scores and identifies where projections diverge from historical baselines.
Current flagship project: AI Infrastructure & Regional Electricity Demand — a causal analysis of data center investment and regional electricity demand growth across PJM, ERCOT, and MISO, using a four-method identification strategy and a forward-looking grid stress forecasting pipeline.