RISHI SIDDHARTH

Data Scientist • Climate Finance & Tech Policy

Bio

I'm a data scientist working at the intersection of climate finance, tech policy, and responsible AI deployment. Currently, I'm building infrastructure at the London School of Economics' Transition Pathway Initiative (TPI) Centre, where I develop APIs that serve climate assessment data to investors managing $87 trillion in assets.

My workridges technical execution and policy understanding. At Georgetown's NSF-sponsored research program, I developed NLP models to assess geopolitical risk from legislative text—work that was presented to the intelligence community. At Intermezzo.ai, I fine-tuned LLMs for international tax compliance, achieving 99% accuracy in entity recognition across OECD, HMRC, and Singapore CPF frameworks.

I'm particularly interested in stablecoin regulation, AI governance, and how data science can inform better policy decisions. I hold a BS in Data Science with an International Business minor from American University and the London School of Economics. I'm a U.S. citizen and security clearance eligible.

Featured Projects

TPI Climate Assessment API

Transition Pathway Initiative - London School of Economics | Jan 2025 - April 2025

Co-developed FastAPI application providing programmatic access to TPI Centre's climate assessments covering 2,000+ publicly listed companies and 85 sovereign entities, supporting 156 global investors managing $87 trillion AUM. Built RESTful endpoints for Management Quality (23 indicators) and Carbon Performance data, implementing Pydantic validation and generating carbon-intensity projections against 1.5°C sectoral benchmarks. Integrated ASCOR sovereign assessment framework (covering 85% of global GHG emissions and 90% of global GDP) with Model Context Protocol (MCP) and RAG features for intelligent data querying. Collaborated with 15-member DS205 team under Dr. Jon Cardoso-Silva and TPI Centre staff, applying test-driven development and CI/CD pipelines to deliver production-grade codebase serving real-world investor use cases.

Economic Classification Research

London School of Economics | Certified Research

Published 72-page research paper analyzing economic classification methodologies using 7 thematic dimensions beyond GNI per capita. Developed multidimensional development index incorporating income, infrastructure, trade, human capital, environmental, institutional, and health factors. Revealed policy-relevant distinctions obscured by traditional GNI classification—identifying middle-income countries with robust human capital but weak infrastructure, and countries with similar GNI but vastly different development profiles. Collaboration with LSE Data Science Society and economics students using World Bank API data, Python, and advanced statistical modeling.

Global AI Adoption Analysis

Independent Research | 2024

Analyzed AI usage patterns across 200+ countries using Anthropic's Economic Index. Found emerging markets show 4-6 percentage point higher automation rates than developed markets—challenging assumptions about AI adoption being concentrated in wealthy nations. Published comprehensive research report examining implications for global development, workforce transformation, and policy responses. Utilized Python for data analysis and visualization, creating interactive dashboards toxplore cross-country patterns.

Stablecoin Policy Research

In Progress | 2026

Developing comprehensive policy analysis of stablecoin regulation, focusing on the GENIUS Act and Treasury Department frameworks. Examining AML/KYC requirements, reserve transparency, systemic risk implications, and international regulatory coordination. Research bridges technical understanding of stablecoin mechanisms with policy implementation challenges, drawing on insights from TPI climate finance work and Intermezzo.ai compliance automation experience.

Other Work

E-Commerce Listing Classifier

Random Forest classifier for product quality assessment. 98% accuracy across 9K product titles with critical analysis of data leakage limitations.

War Monitor (Ukraine)

Multi-agent AI system combining OSINT data streams including ACLED, satellite imagery, government sources, and news using Chain-of-Debate framework.

Economic Inflation Prediction

ML models predicting historical inflation patterns and Bank of England interest rate changes using advanced time series analysis.

Luxembourg Stock Exchange API

Python API integration for extracting and analyzing financial data with automated image processing.

Writing

I write about AI governance, tech policy, climate finance, and data science. My work explores how technical systems intersect with regulatory frameworks and policy decisions.

Read on Substack →