Stefanus Candra
Software Engineer II at Allen Institute for Artificial Intelligence (AI2)
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolStefanus Candra is a Software Engineer II with a decade of experience building production-grade systems and machine learning tooling, currently at the Allen Institute for AI in Seattle. He blends research-caliber ML and computer vision expertise—evidenced by contributions to well-known repos like LPIPS and image colorization projects and a PhD-level background from UC Berkeley—with strong production engineering skills in data pipelines and Apache Spark/Airflow from his time at Whitepages. Comfortable moving between research and engineering, he has a track record of improving model architectures, refactoring complex codebases, and shipping robust data ingestion and publishing systems. Outside of work he’s a recreational musician, reflecting a pragmatic commitment to craft and work-life balance.
10 years of coding experience
5 years of employment as a software developer
B.A, Computer Science, 3.77, B.A, Computer Science, 3.77 at University of California, Berkeley
Spanish, Portuguese, Indonesian
Github Skills (22)
pytorch10
convolutional-neural-networks10
caffe10
python10
image-processing10
machine-learning10
deeplearning-ai10
deep-learning10
deep-neural-networks10
model-optimization10
neural-network10
computer-vision10
colorize9
tensorflow9
trainings9
Programming languages (2)
Github contributions (5)
richzhang/colorization
Mar 2016 - Oct 2020
Automatic colorization using deep neural networks. "Colorful Image Colorization." In ECCV, 2016.
Role in this project:
ML Engineer
ML EngineerContributions:163 commits, 3 PRs, 164 pushes in 4 years 7 months
Contributions summary:Stefanus primarily updated and expanded the demo code for image colorization using deep neural networks. Their contributions involved modifying the demonstration notebooks, adding code related to model loading, image preprocessing, and running the colorization process, as well as adding training code and feature learning models. The user also added a script (`colorize.py`) to enable colorization using command-line arguments.
image-colorizationdeep-learningeccvcomputer-visionneural-networks
richzhang/PerceptualSimilarity
Jan 2018 - Oct 2021
LPIPS metric. pip install lpips
Role in this project:ML Engineer
ML EngineerContributions:2 releases, 115 commits, 7 PRs in 3 years 9 months
Contributions summary:Stefanus primarily contributed to the development and refinement of a perceptual similarity metric, LPIPS. Their work involved modifications to the core distance model, including upsampling and spatial adjustments. The user also implemented a perceptual loss class and integrated it with training code, demonstrating a focus on deep learning model development and experimentation. Further contributions added to the training process and bug fixes.
pytorchpythonpipperceptualdeep-learning
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.