Chanwut Kittivorawong
Graduate Student Researcher
Berkeley, California, United States
Summary
🤩
Rockstar🎓
Top SchoolChanwut Kittivorawong is a PhD student and graduate researcher at UC Berkeley specializing in efficient video analytics and ML inference systems that exploit spatiotemporal and geospatial redundancy to reduce compute cost. He designed Spatialyze, a VLDB 2024 paper project that balances model accuracy and performance via proxy models and input compression, and has practical experience optimizing large-scale ML training and inference pipelines. His internships at Google (video QA for Gemini), OctoML (web-based model visualizers using D3/TypeScript and Vega), and DocuSign (cost-saving AWS automation and Spark jobs) reflect a blend of systems engineering and applied ML. Chanwut is fluent in Python and TypeScript, contributes to notable open-source visualization projects like Vega/Vega-Lite (label/layout and composite mark improvements), and brings eight years of engineering experience. He combines rigorous research with production-minded engineering, particularly around cost-aware inference trade-offs. Based in Berkeley, he is seeking internships in ML systems and large-scale ML acceleration and infrastructure.
8 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of California, BerkeleyMaster of Science - MS, Computer Science and Engineering, Master of Science - MS, Computer Science and Engineering at University of Washington
German, Thai, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (14)
data-visualizations10
typescript10
data-visualization10
javascript10
typescripts10
visualization10
data-visualisation10
visualizations10
vega-lite10
vega10
typescript-types10
charts10
d3js9
canvas9
Programming languages (8)
Github contributions (5)
vega/vega-lite
Jan 2018 - Jan 2022
A concise grammar of interactive graphics, built on Vega.
Role in this project:
Full-stack Developer
Full-stack DeveloperContributions:33 reviews, 75 commits, 60 PRs in 4 years
Contributions summary:Chanwut primarily contributed to the development and refactoring of the Vega-Lite library's composite marks, specifically focusing on error bars and error bands. They implemented new features such as error band support, including interpolation and tension properties, and refactored existing code for box plots, error bars, and tooltip functionality. Additionally, the user addressed issues related to axis customization and overall code structure, demonstrating a strong understanding of the library's architecture and visual analysis principles.
chartsinteractive-graphicsvisual-analysisplotgrammar
vega/vega
Mar 2019 - Jan 2022
A visualization grammar.
Role in this project:Full-stack Developer
Full-stack DeveloperContributions:1 review, 42 commits, 17 PRs in 2 years 10 months
Contributions summary:Chanwut primarily contributed to the `vega/vega` project by implementing and improving the label transform functionality, a core feature of the visualization grammar. Their work included adding new label layout methods, fixing bugs related to label collisions and boundary issues, and enhancing the typing of label-related properties. The user also added tests for the label transform and addressed issues arising from empty data, thereby improving the stability and functionality of the project.
svgcanvasgrammarvegavisualization-grammar
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.