Hi, I’m Sijin Yang

I turn messy systems into clearer decisions.

I’m a University of Washington Informatics student working across product, data, strategy, and engineering. I like ambiguous problems where the answer is not obvious yet — whether that means designing an AI product, evaluating a trading hypothesis, studying community behavior, or improving a real operating system.

<aside> 💡

Currently focused on internships in product, technical PM, consulting, data, and technology.

</aside>

Quick links

Resume · LinkedIn · GitHub · [email protected]

Selected work

📊 NewsTrader — Trading Research Platform

Product • Data • Full-Stack Engineering • Experimentation

I built a paper-only algorithmic trading research platform around a harder question than “can I make a strategy look profitable?” — can the system tell me when my own idea should not be trusted?

• separated historical, prospective shadow, and paper-execution evidence;
• added data-quality and point-in-time safeguards;
• designed human review gates and operator tooling;
• kept live execution disabled and automatic promotion out of the workflow;
• increased historical evidence-collection throughput from about 1.78 → 3.49 selections/session in one experimental replay, without presenting it as a profitability result.

NewsTrader — Trading Research Platform

GitHub repository

📚 Course Companion — Context-Aware AI Tutor