About

I've been the AI person teams bring in before they have an AI team — six years across surgical robotics, finance, and the enterprise agent platform I now lead.

Agentic engineering is a commodity. What isn't: choosing what to build, where automation stops, and whether thousands of people can rely on it. I'm business focused — I construct architectures that last and kill work that isn't earning its keep.

I build the team as much as the system — shared design practice, review conventions, an AI bar the whole team works to, not a personal hack.

Some of this work predates agentic AI as a category — embodied navigation at UCLA, surgical prototypes at J&J, portfolio automation measured against P&L. Same problem, different names.

Off the clock: strategy games, roguelikes, board games — systems that reward long-term thinking. Former board game club president.

Based in Taipei. Working across Silicon Valley, NYC, and Asia — one location, continuously global.

Applied AI

Agentic systems Multi-agent orchestration LLM evaluation MCP Reinforcement learning

Languages

Python C++ JavaScript

Cloud & delivery

GCP Terraform AWS Docker

Domains

Production AI Surgical robotics Portfolio automation

2014 — 2019

Doctor of Philosophy (PhD), Robotics and AI — ABD

University of California, Los Angeles

Coursework and research complete.

Communicative Learning with Natural Gestures for Embodied Navigation Agents with Human-in-the-Scene, Wu et al., IROS 2021

2014 — 2016

Master of Science (MS), Robotics

University of California, Los Angeles

2010 — 2014

Bachelor's degree, Mechanical Engineering

National Taiwan University

Presidential Award.

"Jimmy taught me how to view technical problems from perspectives of engineers and business leaders alike."

Brooke Towns

R&D Leadership Development Program, Johnson & Johnson

"With his rich background in robotics, AI, software engineering, and automation, Jimmy is a rare asset."