Sophomore at Yale, studying cognitive science. Building tools for biotech research.
I'm interested in how decisions get made — especially in markets, where the gap between rational behavior and observed behavior is often where the interesting questions live. Most of what I build sits at the intersection of cognitive science, financial data, and applied AI.
Currently spending the summer at a biotech-focused hedge fund, building internal research tools. Originally from Hong Kong.
Building an earnings call language analyzer for biotech names — looking for patterns in management hedging, confidence shifts, and tone deltas across quarters. Writing up the first findings.
Ingests 13F filings from biotech-focused hedge funds, parses positioning data, and surfaces consensus holdings, exit positions, and concentration changes — collapsing what was a multi-hour manual workflow into a single command. Built with Claude Code.
Analyzes biotech earnings call transcripts for hedging patterns, confidence signals, and tone shifts quarter over quarter. Cross-references against clinical catalysts to surface management language deltas around upcoming readouts.
Autonomous research agent that takes a ticker and produces a structured research memo by chaining 13F positioning, earnings language analysis, clinical trial data, and FDA calendar events. The goal: five minutes for what currently takes four hours.