Can Balioglu
Research Engineer At FAIR at Meta
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolCan Balioglu is a research engineer at Meta FAIR with 11 years of experience building scalable ML and distributed systems. His career spans senior engineering roles at PyTorch, Amazon SageMaker and Alexa AI, plus an entrepreneurial stint co-founding ALK Solutions, giving him both product and infra perspectives. He’s an active contributor to PyTorch—improving CUDA memory handling and DistributedDataParallel robustness—so his work directly impacts one of the most widely used deep learning frameworks. Based in New York, he combines low-level GPU/debugging expertise with production distributed training experience, often focusing on reliability and maintainability improvements that quietly reduce production incidents.
11 years of coding experience
15 years of employment as a software developer
High School, High School at Istanbul Erkek LisesiDiplom, Computer Science, Diplom, Computer Science at Technical University of Munich
Github Skills (12)
cuda10
pytorch10
distributed-training10
gpu10
python10
machine-learning9
deep-learning9
tensor9
neural-network8
autograd8
cprogramming-language7
c-language7
Programming languages (6)
Github contributions (5)
pytorch/pytorch
Feb 2021 - Jul 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:257 reviews, 153 commits, 122 PRs in 1 year 5 months
Contributions summary:Can's contributions focus on improving the PyTorch library's CUDA functionality and distributed training capabilities. They addressed issues related to CUDA memory management, specifically regarding host memory pointers and the `cudaPointerGetAttributes()` function, improving the reliability of CUDA device detection. Additionally, the user worked on the distributed training aspect of PyTorch, fixing bugs in `no_sync` operations when using the `find_unused_parameters` feature in DistributedDataParallel. Finally, the user made improvements to the code's readability and maintainability.
pythongpu-accelerationdeep-learninggpunumpy
facebookresearch/fairseq2
Aug 2023 - Apr 2025
FAIR Sequence Modeling Toolkit 2
Contributions:4 releases, 325 reviews, 1523 PRs in 1 year 7 months
artificial-intelligencedeep-learningmachine-learningpythonpytorch
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.