Richard Zamora
Senior Systems Software Engineer at NVIDIA
Naperville, Illinois, United States
Summary
🤩
Rockstar🎓
Top SchoolRichard Zamora is a Senior Systems Software Engineer with nine years of experience building high-performance data infrastructure, currently driving GPU-accelerated I/O and distributed data processing at NVIDIA on the RAPIDS and NVTabular stacks. He blends HPC and production engineering expertise from national labs with practical open-source contributions to prominent projects like cuDF, Dask, and fsspec, where he has optimized Parquet I/O, byte-range merging, and large-scale shuffle/partitioning. His work shows a pattern of turning deep systems knowledge into measurable performance gains—e.g., refactoring shuffle to use host memory and enabling multi-file partition aggregation for remote storage. With a PhD-trained background in computational materials and exascale I/O practices, he uniquely bridges scientific computing rigor and scalable ML data engineering.
9 years of coding experience
8 years of employment as a software developer
M.S., B.S., Structural Engineering, M.S., B.S., Structural Engineering at Cornell University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (22)
filesystem10
preprocessing10
python10
pyarrow10
preprocess10
data-engineering10
cudf10
feature-engineering10
fsspec10
dask10
gpu10
performance-optimization10
parquet10
data-structure9
c-language9
Programming languages (10)
Github contributions (5)
dask/dask
May 2022 - Jan 2023
Parallel computing with task scheduling
Role in this project:
Back-end Developer & Data Engineer
Back-end Developer & Data EngineerContributions:893 reviews, 4 commits, 337 PRs in 8 months
Contributions summary:Richard primarily contributed to the improvement of the Dask-Dataframe library, focusing on supporting and optimizing Parquet file handling. Their work included adding a feature to specify task partition sizes, enabling filtering based on partitioned columns, and resolving issues related to handling the schema of Parquet files. The user also refactored parts of the Parquet API, addressing issues like string-based index handling and ensuring compatibility with newer versions of PyArrow.
pythonschedulingparallelnumpydask
rapidsai/cudf
May 2019 - Jan 2023
cuDF - GPU DataFrame Library
Role in this project:Back-end Developer
Back-end DeveloperContributions:669 reviews, 658 commits, 256 PRs in 3 years 8 months
Contributions summary:Richard's primary contribution appears to be focused on enhancing and optimizing the cuDF library, particularly regarding reading and writing Parquet files. They significantly enhanced the performance of the I/O operations related to remote storage by leveraging fsspec and PyArrow. Additionally, the user implemented features such as multi-file partition aggregation and fixed a bug in the handling of index columns in Parquet files.
cudadataframe-librarydata-analysiscppcudf
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.