End-to-end Machine Learning

I’m an applied statistician and machine learning engineer. I work on two problems that resist clean measurement, climate change and illicit finance, and on the methods that make them tractable. I co-founded Plan4Planet, a community that connects AI planning and reasoning researchers with climate experts. I developed the United Nations Economic Commission for Africa’s official methodology for estimating trade misinvoicing across 167 countries, and the risk-based framework I authored underpins the Tax Justice Network’s Illicit Financial Flows Vulnerability Tracker. I specialize in building rigorous, end-to-end AI systems for high-impact domains, from climate finance to financial crime, and in the question underneath all of them: how to evaluate a model, a market, or a policy when the ground truth is hidden. With a PhD in environmental science and a master’s in statistics, I bring both technical rigor and domain fluency to problems that cut across disciplines.

I’m a full-stack developer: I like building things that work and shipping them.

My work leverages machine learning and econometric techniques to address two critical global challenges: climate change and illicit finance. By training, I am a political economist and statistician: I use innovative quantitative techniques and methods from statistical data science to answer questions about pressing socio-economic problems. I was most recently a Postdoctoral Fellow with the Climate Solutions Lab and the Rhodes Center for International Economics and Finance at the Watson Institute for International and Public Affairs at Brown University, where I concentrated on how to catalyze private sector investment for a greener future.

My research explores the interplay between money, markets, and public policy, whether it's paying for climate action or the shadowy realm of illicit economies. In the area of climate finance, I analyze market-driven solutions to climate change, such as carbon taxes and novel financing mechanisms. Conversely, my work on illicit finance reveals the darker side of markets, shedding light on the impact of money laundering and financial secrecy on the sustainable development of emerging economies. My work in this space is similar to investigative statistics: it involves finding statistical evidence of illicit activity in economic data, behavior that, by definition, seeks to remain hidden.

I am a leader in the emerging field of climate change and artificial intelligence: contributing to sustainability projects in climate tech as a lead machine learning engineer, speaking on the role of AI in the low-carbon economy, and serving as a jury member for various AI competitions.

With a track record of publications in scientific journals and high-level political advisory work, my contributions bridge the gap between research and policy. My specialty is combining quantitative research with pragmatic policy proposals and designing intuitive tools to aid decision-making. My work has been sought after by international organizations including the United Nations, the Tax Justice Network, and the Business & Sustainable Development Commission. For the United Nations Economic Commission for Africa, I spearheaded their estimates of illicit trade, a key component in understanding the financial outflows from the continent.

I obtained my PhD in Environmental Science and Management at the University of California Santa Barbara, where I developed a methodological toolkit to analyze complex policy problems. While I was at UCSB, I was affiliated with the ENVENT and the Conservation Economics labs.

Prior to my PhD, I was a researcher at the leading development economics think tank Center for Global Development for several years, where I built and developed a simulation tool for policy-makers, available as an interactive website at SkyShares, which computes the environmental and economic costs of various climate agreements. I have also worked for the African Secretariat of ICLEI – Local Governments for Sustainability in Cape Town.

Interdisciplinarity and multiculturalism are the defining features of my career. My training spans various disciplines, having obtained an MA in Statistics from UCSB, an MSc in Economic History from the London School of Economics, and an MSc in Economics and Public Policy from Sciences Po Paris and École Polytechnique. I earned a First Class Honours BA in European Social and Political Studies from University College London. This educational journey mirrors my international upbringing in Taiwan, Abu Dhabi, and Spain.

Fluent in three languages (and several programming ones), I distinguish myself by contributing to various disciplines at the highest level, both in academia and policy, and by my ability to make complex topics accessible to diverse audiences. I excel at translating concepts and ideas across topics to engage and connect with diverse groups. My teaching stints at UCSB were particularly rewarding, receiving enthusiastic feedback.

A cornerstone of my career is integrating methodological rigor and highly technical skills with deep substantive expertise in environmental policy and development finance.

I thrive in dynamic environments, having played a key role in establishing the European office of the Center for Global Development. I'm insatiably curious and continually seek to learn new skills. I like web design, training my German Shepherd Oso, and indoor gardening (my portfolio may not reflect all of those interests!).

Founding

Planning for a Better Planet

I co-founded Plan4Planet, a community that connects AI planning and reasoning researchers with climate experts, so that AI research is driven by real climate problems. Our inaugural event at the 2026 AAAI Fall Symposium series brought together speakers from the World Bank, Cornell, the University of New Hampshire, and the University of Maryland, 14 accepted papers, and three days of working sessions. I opened it with the framing talk on the normative choices hidden in planning for climate. The community continues through a steering committee and working groups.

Visit plan4planet.org

Featured course

10 self-paced modules

Free course! Build the skills to assess evidence, question news headlines, and unpack the logic behind claims in public policy, economics, and the social sciences. No prior experience with statistics required.

Start Learning

New research!

Partisanship and perceived costs predict carbon pricing opposition better than objective costs

Out now in Climatic Change (2026), with Matto Mildenberger, Kathryn Harrison, Chloé Boutron, and Erick Lachapelle. Drawing on a panel survey across five Canadian provinces, we show that perceived price changes explain opposition to carbon taxes better than actual ones, and that partisanship shapes the perception. We tell the story for a general audience in The Conversation: Misinformation won: How divisive politics doomed Canada’s carbon tax.

Read the paper