Zehao Shi
Boston, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolZehao Shi is a Data Science undergraduate at Northeastern (class of 2026) with a solid foundation in mathematics and practical experience applying statistical and Python-based analysis to real-world problems. He has hands-on skills in Linux system management, virtualization (VMware, Parallels, TeamViewer), R and Python (Pandas, NumPy), and LaTeX for reproducible research and reporting. Zehao led a team analyzing county-level business performance during COVID-19, visualizing results for 700+ counties and demonstrating strength in end-to-end data projects. He also contributed as an ML engineer to the MXNet Mask R-CNN repository, improving mask prediction robustness and ROIAlign performance—an uncommon open-source footprint for a current student. Recent industry exposure includes a commercial real-estate data internship at JLL Technologies, where he collected and synthesized data for investor consulting. Curious and detail-oriented, he combines systems-level tooling familiarity with statistical rigor to bridge data engineering and applied analysis.
10 years of coding experience
NortheasternPinghe High School
Chinese, English
Stackoverflow
99reputation
3kreached
8answers
7questions
Github Skills (11)
mask-rcnn10
mxnet10
computer-vision10
machine-learning10
python10
object-detection6
ios6
deep-learning6
matlab6
tensorflow6
cascade-classifier6
Programming languages (5)
Github contributions (5)
TuSimple/mx-maskrcnn
Oct 2017 - Nov 2017
An MXNet implementation of Mask R-CNN
Role in this project:
ML Engineer
ML EngineerContributions:22 commits, 16 PRs, 11 pushes in 1 month
Contributions summary:Zehao primarily focused on improving and fixing issues within the Mask R-CNN implementation. Their contributions include fixing array boundary risks and missing configurations related to training scales. They also improved mask metrics, modified weights of the mask branch, and addressed issues related to mask flipping and ROIAlign optimization. The commits demonstrate a focus on the core functionality and performance of the mask prediction module.
pytorchmaskmxnetdeep-learningr-cnn
Zehaos/TFSTN
Apr 2017 - Apr 2017
Contributions:24 commits, 25 pushes, 1 branch in 7 days
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.