G Ramalingam
Principal Researcher at Microsoft
Bellevue, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolG Ramalingam is a Principal Researcher at Microsoft with over three decades of academic and industrial experience and more than 10 years focused on modern ML engineering and compiler back-ends. He holds a PhD in Computer Science and dual MS degrees from the University of Wisconsin–Madison, and a BTech from IIT Madras, reflecting deep theoretical and practical foundations. At Microsoft and through long tenure at IBM, he has specialized in model representation, type/shape inference, and runtime optimizations for ML systems. His open-source contributions include substantial work on ONNX and ONNX Runtime—adding sparse tensor operators and improving model lowering in ONNX-MLIR—demonstrating an ability to bridge compiler infrastructure and production inference engines. Colleagues value his meticulous approach to correctness, shown by extensive type/shape refinements and comprehensive test coverage. Based in Bellevue, WA, he combines research rigor with pragmatic engineering to make ML runtimes more robust and efficient.
10 years of coding experience
4 years of employment as a software developer
Indian Institute of Technology MadrasMaster of Science - MS, Computer Science, Master of Science - MS, Computer Science at University of Wisconsin-Madison
Github Skills (20)
c-language10
python10
sparse-matrix10
machine-learning10
inference10
onnx10
mlr10
compiler-design10
deep-learning10
sparse-data10
cprogramming-language10
shapes10
type-inference9
computer-engineering8
tensorflow8
Programming languages (7)
Github contributions (5)
onnx/onnx
Sep 2017 - Jan 2023
Open standard for machine learning interoperability
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1960 reviews, 84 commits, 494 PRs in 5 years 5 months
Contributions summary:G's commits primarily focused on modifying the ONNX definition files, particularly the .proto and .cc files, indicating a role in enhancing the ONNX framework itself. They updated the type and shape inference functions for multiple operations including basic math, control flow, and experimental operators. This involves deep understanding of the framework internals. They also contributed to test cases, which points towards a role of back-end development and also demonstrates ML Engineering skills.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
microsoft/onnxruntime
Jun 2019 - Aug 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:ML Engineer
ML EngineerContributions:200 reviews, 52 commits, 45 PRs in 3 years 3 months
Contributions summary:G contributed significantly to the implementation of sparse tensor support within the ONNX Runtime. Their work included the creation of new operators like `SparseFromCOO`, `SparseAbs`, and `SparseToValues`, demonstrating a focus on extending the framework's capabilities in handling sparse data structures. The user also addressed build errors, compiler warnings, and PR comments, indicating involvement in code quality and integration within the larger project. The addition of test cases showcases a commitment to ensuring the correctness of their implemented features.
runtimetrainingtensorflowai-frameworkaccelerator
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.