Jaemin Shin
Applied Scientist at Amazon
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolJaemin Shin is an Applied Scientist at Amazon with five years of experience building production-grade ML and federated learning systems, combining deep academic research from KAIST (PhD) with hands-on engineering in industry. He has driven efficient on-device LLM fine-tuning and multimodal mental-health monitoring that improved AUROC and MAE versus prior work, and developed low-power sensing hardware and federated methods recognized at CHI and EMNLP. His internships at Microsoft and Cisco produced measurable speed and accuracy gains in FL systems and led to patent filings and open-source contributions. As a contributor to the widely used dcm2niix project, he improved DICOM→NIfTI conversion for GE scanners, surfacing subtle metadata fixes that benefit neuroimaging workflows. Based in the San Francisco Bay Area, he blends systems-level optimization with applied ML research to move prototypes into scalable, device-aware production.
5 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at 한국과학기술원(KAIST)
Github Skills (7)
c-language10
dicom10
nifti10
cprogramming-language10
neuroimaging10
mri10
python5
Programming languages (4)
Github contributions (5)
rordenlab/dcm2niix
Oct 2020 - Nov 2022
dcm2nii DICOM to NIfTI converter: compiled versions available from NITRC
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1 review, 20 commits, 22 PRs in 2 years
Contributions summary:Jaemin primarily focused on enhancing the functionality of the DICOM to NIfTI converter, particularly for GE medical imaging data. Their work involved implementing and refining the detection of GE group delays and slice timing information, including handling multi-phase EPI and diffusion sequences. They also addressed specific issues related to software version detection and Total Readout Time calculation, improving the accuracy of metadata extraction for the conversion process. Furthermore, the user added the capability to handle GE direct fieldmaps.
neurosciencedcmdicom-imagesbids-formatjpeg
mr-jaemin/ge-mri
Jan 2022 - Jun 2022
GE HealthCare MRI research
Contributions:2 reviews, 23 commits, 3 PRs in 5 months
dicomdiffusiontensor
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.