Los Angeles · M.S. Computer Science, USC

Ruoshi Zhu

朱若时 · Ross

Full-stack engineer & creative developer

I build innovative solutions that combine cutting-edge technology with exceptional user experiences. Currently pursuing a Master’s in Computer Science at USC.

Studying
M.S. Computer Science · USC
Previously
AI/ML engineer · Vensora
Building
Midas, VRC Free Watch, OwlSplitter
Based in
Los Angeles, CA
01 — Work

Selected work

5 projects across web, mobile, desktop and games.

GitHub
More on GitHubgithub.com/54dbd ↗
02 — About

Education & research

Education

University of Southern California
M.S. Computer Science
2025 – Present
GPA 3.95 / 4.00
Shanghai University
B.S. Computer Science and Technology
2020 – 2024
GPA 3.55 / 4.00

Research

01
Quantum Algorithm for Bitcoin Mining
Led a research team studying how quantum technology could affect Bitcoin mining.
02
AI for COVID-19 Diagnosis
CT scan analysis with StyleGAN3 and Swin-UNet.
03
CBA Game Prediction
An STGCN framework for predicting basketball game outcomes.
03 — Experience

Industry

Vensora Inc.
AI/ML Engineer
Dec 2025 – May 2026
LLM AgentsRetrievalCode GenerationEvaluation
Agent-based LLM framework

Built an agent-based LLM framework combining retrieval, execution and automated validation into a code generation pipeline.

Self-correcting feedback loops

Designed feedback loops that use execution results and error analysis to iteratively improve generated code quality.

Workflow orchestration

Optimized LLM workflows for robustness and developer productivity through automated orchestration and evaluation.

Shanghai Shoukou Technology Ltd.
Algorithm Engineer Intern
Aug – Dec 2023
NLPAIOCRAlgorithm
KeYan Reader

AI-powered article reading platform. Worked on NLP and summarization algorithms, improving accuracy and latency.

Cross-team collaboration & support

Deployment, server operations, testing and after-sales technical support.

SoMark

OCR tool that converts PDFs and images to Markdown or JSON for large-scale document digitization.

04 — Publications

Research & publications

ICIAAI 2023 proceedings

Research on Lung Diagnosis Methods based on Data Augmentation

Uses Swin-Unet with StyleGAN3-based dataset expansion for lung diagnosis.

Machine LearningComputer VisionMedical AIStyleGAN3
DOI 10.2991/978-94-6463-300-9_55View paper ↗
05 — Skills

Skills & technologies

Programming

C
C++
Python
JavaScript
TypeScript
Java
Swift

Web development

HTML
CSS
React
Next.js
Node.js
MySQL

AI & ML

TensorFlow
PyTorch
Computer Vision
Deep Learning

Design tools

Photoshop
Illustrator
Premiere
DaVinci Resolve
06 — Contact

Let’s connect.

Email is the best way to reach me.

[email protected]