Experience

My work sits at the intersection of economics, machine learning, and decision-making. I moved from academic research in macroeconomics and firm dynamics into applied science roles focused on experimentation, forecasting, and high-stakes product and business decisions.

Recent Roles

Applied Scientist, LinkedIn

  • Design AI-driven causal reasoning systems that help product and business users translate natural-language questions into credible causal inference workflows
  • Build long-term experiment impact prediction frameworks using surrogate-index methods, Double Machine Learning, and backtesting against matured experiments
  • Combine experimental, observational, and product-behavior data to improve launch decisions when long-term outcomes are delayed
  • Partner with product, engineering, and business teams to turn causal methods into interpretable decision systems

Data Scientist / Applied Scientist, Amazon

  • Led forecasting and causal inference systems for seller growth, shipping, logistics, and operational planning decisions
  • Designed modular forecasting architecture, reactive nowcasting systems, and LLM-assisted forecasting pipeline orchestration
  • Built decision frameworks for forecast-to-cost trade-offs, forecastability diagnostics, market-signal analysis, and external event studies
  • Translated technical methods into reusable tools, internal technical papers, and planning workflows used by business, finance, product, and operations partners

Earlier Roles

Quantitative Researcher, Geode Capital Management

  • Conducted equity alpha research on ESG and sustainability factors, linking granular vendor datasets to testable investment hypotheses
  • Built and backtested proprietary ESG signal variants using multiple structured and alternative data sources
  • Evaluated signal quality through cross-sectional tests, portfolio backtests, robustness checks, and sensitivity to sector or industry exposure
  • Translated noisy ESG inputs into research-ready features and documented where signals appeared economically meaningful versus unstable

Research Associate, Boston University

  • Supported Boston University research connected to NBER-affiliated joint work on political risk, country risk, capital flows, automation, and market power
  • Built research pipelines for large-scale financial text and structured data, including earnings call transcripts, SEC 10-K filings, financial statements, and Compustat/Capital IQ-style fundamentals
  • Applied computational linguistics, feature engineering, and statistical modeling in Python, R, and Stata to construct and validate text-based risk measures
  • Automated figures, tables, text snippets, and model outputs so collaborators could inspect empirical results, refine specifications, and move research drafts forward

Teaching Fellow, Boston University, Universitat Pompeu Fabra, and Barcelona Graduate School of Economics

  • Taught graduate and undergraduate macroeconomics, microeconomics, and quantitative methods
  • Supported course delivery, student learning, and assessment across multiple economics programs

Academic Background

  • Ph.D. in Economics, Boston University
  • Earlier research focused on firm dynamics, macroeconomics, and related quantitative methods

Core Strengths

  • AI agents and reasoning layers for causal decision support
  • Causal inference with experimental + observational data
  • Long-term experiment evaluation under delayed feedback
  • Forecasting infrastructure design, orchestration, and model integration
  • Decision-focused communication with technical and business stakeholders