I am building a physics-centered data practice: learning superconducting quantum systems while developing the tools to generate, manage, analyze, and verify research data.
I am drawn to problems that begin as a rough physical question rather than a clean assignment. My habit is to slow the problem down, name the assumptions, and build a small loop that can be checked by code, simulation, or data.
The road I am building is not about becoming someone who only knows how to use tools. I want tools to become leverage for understanding physical questions; data generation, management, and analysis are ways to make research inspectable, reproducible, and extensible.
AI is part of that loop, but not a replacement for judgment. I use it to widen the search space, draft competing implementations, find blind spots, and accelerate documentation; the final answer still has to survive derivations, numerical checks, and direct inspection.
I care about quiet, reliable systems. A good script, storage setup, simulation report, or note should make the next question easier for future me and for the people working with me.
Selected Work
Floating-Qubit Energy Relaxation Study
Undergraduate thesis, National Central University
Model validation
Admittance formulation
Full-wave FEM
Framed the physical model: Applied an admittance-based formulation to describe floating-qubit coupling paths and make the energy-relaxation mechanism explicit enough for analytical and numerical review.
Cross-checked assumptions: Compared quasi-static extraction with full-wave finite-element simulation to identify where model assumptions stayed consistent and where numerical differences could arise.
Layout-to-FEM Research Workflow
Research data generation through simulation automation
AI-assisted tooling
Simulation automation
Parameter sweep
Built an iteration loop: Developed a data-generation workflow from GDSII layout to FEM simulation, connected to parametric design tooling so layout changes could become solver-ready experiments more quickly.
Used AI as engineering leverage: Worked in human-led, AI-assisted loops to draft automation paths, compare implementation options, review generated code, and keep the tooling tied to the underlying physics constraints.
Mapped circuit parameters: Linked lumped circuit models to Hamiltonian-level quantities around coupling strength and ZZ interaction space, enabling targeted scans for calibration-sensitive parameters.
Compute and Mesh-Refinement Operations
Simulation data analysis and verification
HPC operations
Mesh refinement
Numerical reliability
Ran larger solver workloads: Used National Center for High-performance Computing (NCHC) cluster resources to produce and inspect larger simulation datasets for complex layout FEM models.
Balanced time and accuracy: Tuned mesh-refinement parameters systematically so runtime, resource usage, numerical differences, and reliability could be evaluated with concrete evidence.
Research Infrastructure and Data Environment
Research data management and compute operations
NAS operations
Docker
Research systems
Supported lab data practices: Helped maintain a lab-scale Synology NAS environment, improving backup habits, team access, and day-to-day reliability for research data.
Built personal research infrastructure: Maintained a customized TrueNAS system integrated with Docker for research-data organization, automated backups, and flexible local compute environments.
Education
National Central University
B.S. in Physics · Expected 2026
Thesis: Study of XY-Line External Coupling Effects on Floating Qubit Energy Relaxation