Summary
James Pak is a Staff Machine Learning Engineer with 12 years of experience building search, recommender, and ads ranking systems at scale across Coupang, TikTok, Microsoft, and IBM. He specializes in state-of-the-art ranking architectures, embedding initiatives (including LLM-driven embeddings), and rigorous experimentation—leading 80+ A/B tests and multiple launches that moved core metrics by double-digit percentages. His work spans the full ML lifecycle from backbone and data scaling to online/offline metric ownership and real-time embedding serving for monetization products. Prior roles include productionizing GPT-fine-tuned models for feed personalization and designing position de-biasing and Thompson Sampling exploration methods to improve diversity and partner health. Based in Seattle, he blends research-driven model innovation (KDD submission-level work) with pragmatic systems engineering for production reliability. An early-career systems background and open-source tooling experience give him a rare combination of low-level performance optimization and high-impact ML product delivery.
12 years of coding experience
10 years of employment as a software developer
Master's degree, Computer Science, Master's degree, Computer Science at University of Southern California
Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at Inha University 인하대학교
Certificate, English Language Program, Certificate, English Language Program at University of Pennsylvania
Korean, English