A
Andrew Huang
黃聖瀚
I build healthcare AI that ships — into real hospitals, used by real clinicians.
Sixth-year medical student at NYCU, clerking at Taipei Veterans General Hospital. Co-developer of the ASUS Clinical AI Assistant, deployed to VGHTPE on 2025-11-12. Coming up: NTU Smart MHI master's program this September.
Now shipping
ASUS Clinical AI Assistant智慧病歷助手
VGHTPE AI Impact Center · Clinical–AI Integration Lead & Co-developer
A clinical documentation assistant deployed inside Taipei Veterans General Hospital. Co-developed by bridging clinician workflow, internal engineering and ASUS teams — translating real ward needs into features that survived deployment iteration.
Under Dr. Wei-Chiang Li 李偉強醫師 (北榮副院長) · Dr. Hao-Min Cheng 鄭浩民醫師 (教學部主任) · showcased at Medical Technology Expo (Dec 4–7, 2025)
Selected projects click any card
-

CT Annotation MVP ↗
2026 · open source
117 normal anatomical structures, segmented automatically on any abdominal CT. Whole pipeline runs locally on Apple Silicon — zero cloud, zero PHI risk.
view live demo →「比院內 commercial software 還準。」 — Radiologist + medical physicist, 2026-05 -

Mitral Valve Dynamics ↗
2026 · interactive paper companion
An interactive simulator for mitral valve mechanics — stress distribution, 4D heart visualization, Salgo simulation, FEA comparison. Ten modules walk through every phase of the underlying paper.
explore the simulator → -

Miracle Shiba ↗
Year 2 winter break · founder & builder
An NFT community project built from zero in a single sophomore winter. Concept → community ops → website → on-chain launch — the whole chain.
view on OpenSea → project site ↗
Research in progress
About
Born in Sabah, Malaysia. Schooled at 亞庇建國中學, then Taiwan via NTNU overseas prep, then NYCU School of Medicine in 2020. Primary clerkship at Taipei Veterans General Hospital. This September I begin NTU Smart MHI master's program.
I build because clinical workflows have too many seams that AI can sew. Working on the ASUS Clinical AI Assistant deploy at VGHTPE taught me the entire chain — needs → engineering → real ward usage. The CT annotation MVP came out of wondering whether open-source could beat a license fee; turns out it can, when a radiologist tells you so.
Foundation pacing — Y1 NTU AI coursework, Y2 Miracle Shiba launch, Y4 NVIDIA Developer Program, Y5–6 VGHTPE Smart EMR plus several ongoing studies. AI subscriptions run NT$7,500+/month. Every token has to earn its keep.
Want to talk medical AI, open-source imaging, SR/MA topics, or a project I can help with — email me.
Capabilities
Mentors & collaborators
Teaching & speaking
Beyond medicine
Notes coming soon