MSc FinTech candidate at Frankfurt School of Finance & Management. I think about monetary policy, financial markets, and the systems that move capital across borders. Currently at EMPA Data & Management Consulting in Growth & Strategy. Previously at Volkswagen Group China in Corporate Strategy.
Concentration: Central Banking & Financial Regulation
Program hosted by the Centre for Digital Economics
(FS-C4DE), with the concentration directed by the Centre for Central Banking.
Instructed by leading experts including Prof. Dr. Jens Weidmann, Prof. Dr. Emmanuel Mönch, Prof. Dr. Michael
Ehrmann, and Prof. Dr. Co-Pierre Georg.
Grade: 1,6 (Top of Class) · Coursework: Machine Learning, Digital
Assets, Statistics & Econometrics (1,0), Financial Platforms, Monetary Economics, Central Bank Watching
Boutique consultancy specialising in Data Strategy, AI/Data Governance, and Regulatory Compliance. Formulated GTM strategy, product positioning, and pricing for EMPA's proprietary data governance ERP, and facilitated client demos. Built comprehensive financial scenario models (P&L, cash flow, etc.) to support strategic decision-making. Co-designed the EMPA Venture Standard Process for firm-wide venture development.
Acted as the primary operational and strategic support to the VP of Corporate Strategy. Prepared board-ready Top Management Committee materials and data-driven studies on China's EV landscape to inform product and investment decisions. Defined and tracked OKRs for the "NEW AUTO 2030" initiative, and managed the strategic partnership with XPeng through cross-functional coordination. Additionally served as departmental risk manager, overseeing the identification and assessment of operational risks.
A personal RAG study assistant for CFA Level II preparation — built from scratch: PDF ingestion, hybrid keyword + dense retrieval (bge-m3) with cross-encoder reranking, and grounded answers with passage-level citations, powered by DeepSeek. Evaluated on a 40-question golden set: MRR 0.94, hit@1 92%, ~1–2s per query over a 50,000+ passage corpus, in English or Chinese.
Replicates and extends Aruoba & Drechsel (2024) — LLM-based sentiment scoring of FOMC Greenbooks to identify monetary policy shocks. Finds concept selection is incidental: shuffling all 296 concept labels barely moves the fit (R² −0.03), with shock correlations above 0.99; the result holds for a 595-concept LLM-generated list.
Tests whether GDELT news sentiment can improve Chow-Lin temporal disaggregation of Euro Area GDP, extending Mönch & Uhlig (2005). Monthly: text adds nothing beyond industrial production. Weekly: text alone yields a plausible path (R² = 0.94) that reacts correctly to 5/6 major economic events.
Backtest of the pre-FOMC announcement drift documented by Lucca & Moench (2015) — the equity premium earned in the 24 hours before scheduled FOMC announcements.
A web app for tracking fat loss through daily logging of food intake, activity, and weight. Uses AI (DeepSeek V3) to parse natural language meal descriptions into detailed macro breakdowns, with a P&L-style energy accounting system covering BMR, thermic effect, and exercise calories. Includes historical trend charts, multi-user authentication, and CSV export. Accessible at cut.john-huang.com ↗.
An end-to-end anti-money laundering compliance platform integrating real-time transaction screening, entity resolution, and risk-scoring workflows. Designed to help financial institutions streamline SAR filing and regulatory reporting through automated case management and audit-ready documentation.
A regulatory-focused diagnostic tool that quantifies feature redundancy in credit risk models using sparse random projections. Assesses effective dimensionality of credit datasets to identify over-parameterization and ensure compliance with Basel IRB and IFRS 9 validation frameworks. Analyzes the Home Credit Default Risk dataset (300k samples, 59 features) to detect redundancy patterns.
A financial distress prediction model diagnostic using sparse random projections to quantify feature redundancy. Applied to the Polish Companies Bankruptcy dataset to measure intrinsic dimensionality of 64 financial ratio features and validate model complexity. Identifies 21.9% redundancy ratio, enabling feature selection and improved model generalization.