01
Diagnose
Separate the intended capability from the benchmark, proxy, or failure that first exposed it.
/ ABOUT
Yuhua (Bill) Chen.
I turn capability gaps into trainable, verifiable systems.
/ EXPERIENCE
01Amazon AGI — Applied Scientist, 2025–present
02Q.bio — Machine Learning Engineer through Staff Machine Learning Engineer & AI Research Lead, 2021–April 2025
03UCLA — Ph.D
04in Bioengineering, 2016–2021
/ ONE ARC
Across medical imaging and frontier-model post-training, my work has followed the same pattern: identify where the current formulation breaks down, redefine the problem around the real bottleneck, and build the full system needed to test the new formulation in practice.
In medical imaging, the challenge was recovering useful structure from incomplete physical measurements.
In autonomous MRI, models moved inside a live scanner workflow, where inference speed and real-time decisions mattered as much as offline accuracy.
In frontier-model post-training, the work centers on the systems that create and maintain useful learning signals: environments, verifiers, curriculum, rollouts, reinforcement learning, and evaluation.
/ HOW I APPROACH THE WORK
02
Reformulate
Choose a unit of training and verification that makes the real bottleneck observable and changeable.
03
Build and test
Connect research, runtime systems, and evaluation closely enough that the result can revise the next formulation.
/ EARLIER SYSTEMS
At Q.bio, I deployed 3D AI systems inside an autonomous MRI scanner, including components with sub-second inference for real-time localization and imaging decisions.
My earlier work in 3D MRI reconstruction, super-resolution, and segmentation has been cited more than 1,000 times.
I am a co-inventor on patent applications covering deep-learning MRI reconstruction and calcium-free CT angiography.
/ TRAINING
Ph.D. in Bioengineering from UCLA; M.S. degrees in Computer & Information Technology and Bioengineering from the University of Pennsylvania.
Outside work, running is the most durable rhythm in my life. The appeal is simple: steady effort compounds, conditions are never fully controllable, and progress has to survive contact with reality.