Rahul Huilgol
Member Of Technical Staff at Parallel Web Systems
Sunnyvale, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRahul Huilgol is a senior software engineer with 12 years of experience specializing in the intersection of AI and high-performance distributed systems, currently building web search for AI at Parallel Web Systems. He previously led efforts at AWS on the Neuron stack for Trainium and on SageMaker’s Distributed Model Parallel library, developing novel sharding, parallelism and memory-saving strategies to scale large-model training. Rahul has deep hands-on expertise shipping production deep-learning infrastructure—optimizing kernels, collectives, and CI/CD for custom TensorFlow/PyTorch binaries—and contributed to widely used open-source projects like MXNet and AWS Deep Learning Containers. His background spans both research internships (including work with Yoshua Bengio) and production engineering, giving him a rare bridge between ML research and systems engineering. Based in Sunnyvale, he brings a pragmatic focus on performance and reliability, and a proven track record of turning compiler- and build-level fixes into more stable, high-throughput training platforms. An interesting detail: beyond system design he’s directly implemented low-level improvements such as F16C support and build-system fixes that materially improved portability and performance.
12 years of coding experience
9 years of employment as a software developer
Bachelor of Technology Computer Science and Engineering, Bachelor of Technology Computer Science and Engineering at Indian Institute of Technology, GuwahatiSecondary Education, Secondary Education at Sanghamitra SchoolSenior Secondary Education Math Physics Chemistry, Senior Secondary Education Math Physics Chemistry at Sri Chaitanya Junior CollegeMaster of Science (M.S.) Computer Science, Master of Science (M.S.) Computer Science at The University of Texas at AustinStanford Continuing Studies
English, Hindi, Kannada, Telugu
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (25)
pytorch10
docker10
c-language10
compile-time10
compilation10
mxnet10
build-system10
dockers10
sagemaker10
mlops10
tensorflow10
aws10
compile10
compiler10
cprogramming-language10
Programming languages (11)
Github contributions (5)
apache/mxnet
Jul 2017 - Jul 2019
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:
Back-end Developer
Back-end DeveloperContributions:43 commits, 99 PRs, 1 push in 2 years
Contributions summary:Rahul primarily focused on fixing compiler warnings, build errors, and compilation issues within the MXNet codebase, indicating a focus on code quality and stability. Their contributions included adding casts, overrides, and addressing errors related to OpenMP and linting, specifically related to the C++ codebase. Additionally, they worked on resolving issues related to the build process, including Jenkins scripts, and made changes to several core operator files, demonstrating a deep understanding of the project's underlying implementation.
pythonschedulerdataflowmutationdata-science
dmlc/mshadow
Jul 2017 - May 2018
Matrix Shadow:Lightweight CPU/GPU Matrix and Tensor Template Library in C++/CUDA for (Deep) Machine Learning
Role in this project:Backend Developer
Backend DeveloperContributions:5 commits, 4 PRs, 22 comments in 9 months
Contributions summary:Rahul primarily focused on improving the performance and stability of the `mshadow` library. Their contributions included fixing compiler warnings related to sign comparisons, and implementing F16C instruction set architecture extension for the half datatype. They also addressed lint errors and compiler warnings while making changes to `tensor_cpu-inl.h`, `tensor.h`, and `half.h`. They were also involved in updating the build process with CMake.
cudapytorchcpucppdeep-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.