HarryHu
I turn ambiguous real-world problems into deployable ML systems — building computer vision pipelines, multimodal AI applications, and data products that ship with measurable impact.
Curated highlights across work experience and research projects — emphasizing real constraints and measurable impact.
Under-utilized space uncovered
Blueprint processing time cut
Vision pipeline mAP gain
Production ML deployments
Washington University in St. Louis
M.Sc. in Data Analytics & Statistics
McKelvey School of Engineering
Aug 2024 – Dec 2025
New York University
B.Sc. in Data Science & Mathematics
College of Arts & Science
Jan 2021 – May 2024
I'm an AI/ML Engineer and Data Scientist with a background in data science, statistics, machine learning, and applied research. My interests sit at the intersection of modeling, systems thinking, and real-world problem solving — especially in computer vision, multimodal AI, and decision-focused analytics.
I'm most energized by messy, high-ambiguity problems that require more than just fitting a model: defining the right objective, working through constraints, choosing practical tradeoffs, and building toward something that can actually be used.
Projects built entirely with AI agents — Claude Code, OpenClaw, and AI-assisted workflows. Not autocomplete. Collaborative engineering at full velocity.
Personal Portfolio Website
This site — designed and built entirely through AI-agent-augmented development with Claude Code. Interactive canvas background, glassmorphism design system, click-to-expand timelines, and full-stack Next.js deployment.
AI Livestream Video Generator
An AI-powered livestream simulation tool that generates dynamic video content. Built end-to-end with Claude Code from system design to implementation.
Clash Royale Elixir Calculator
A utility app for calculating optimal elixir costs and deck compositions in Clash Royale. Vibe-coded from concept to working app in a single session.
What collaborators, stakeholders, and teammates say about working with me.
Harry exemplified enthusiasm, energy, and real depth from day one. He supported new technologies I launched for space-utilization and analytics, and he approached every challenge with a level of maturity well beyond an internship role. What stood out most was his consistency — he asked thoughtful questions, iterated quickly, and delivered work that directly strengthened our operational decision-making. Harry is a standout data scientist with strong instincts, a collaborative spirit, and a genuine commitment to meaningful work.
Exploring multimodal AI systems, applied computer vision, and production ML workflows — while building case-study-driven projects at the intersection of research and real-world deployment.
Let's build
something useful.
I'm always interested in thoughtful conversations about machine learning, applied AI, computer vision, and building useful systems.