Oliver Chang
Software Engineer Intern at Omni
California, United States
Summary
👤
Senior🎓
Top SchoolOliver Chang is a PhD candidate and graduate student researcher at UC Santa Cruz with eight years of experience building and benchmarking reinforcement learning systems for autonomous driving. He has cut RL training time dramatically through transfer-learning implementations in PyTorch, revealed algorithmic stability differences between SAC and PPO via rigorous simulator benchmarks, and optimized memory use with image-preprocessing to enable large replay buffers. Oliver combines research rigor—with a first-author IEEE conference paper and multidisciplinary lab experience—with production-minded engineering, having led teams to deploy models across multiple simulators using Docker, Kubernetes, and GitHub. He also brings applied AI experience from industry, automating company workflows with Azure Functions and OpenAI APIs to save substantial licensing costs and employee time. Based in California, he balances deep technical chops in RL, CV, and cloud orchestration with off-hours passion for running and baseball, reflecting a methodical but energetic approach to problem solving.
8 years of coding experience
2 years of employment as a software developer
University of California Santa CruzBachelor of Arts - BA, Mathematics and Computer Science, Bachelor of Arts - BA, Mathematics and Computer Science at Pomona College
Spanish
Github Skills (25)
toolbox9
robotics8
simulator7
reinforcement-learning7
stable-baselines7
python7
transfer-learning6
autonomous-driving6
meta-learning6
machine-learning-algorithms6
machine-learning6
pytorch6
openai5
autonomous-vehicles5
system-design5
Programming languages (6)
Github contributions (5)
oliverc1623/DRIVE-Sim
Jun 2023 - May 2024
A PyTorch-based framework to conduct deep reinforcement learning research in multiple autonomous vehicle simulators
Contributions:90 PRs, 415 pushes, 3 branches in 11 months
deep-reinforcement-learningpytorchautonomous-vehiclesneural-networkspolicy-gradient
oliverc1623/luau
Jun 2024 - Aug 2025
This project aims to implement and extend the Introspective Action Advising (IAA) algorithm by developing an adaptive introspection mechanism that dynamically adjusts the relevance of teacher-provided advice based on real-time feedback in target domains.
Contributions:10 PRs, 495 pushes, 13 branches in 1 year 2 months
domain-adaptationinterpretable-machine-learningmeta-learningppopytorch
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.