Shubham Bhokare
Software Engineer at Microsoft
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolShubham Bhokare is a software engineer with eight years of experience building AI platforms and ML tooling, currently contributing to Microsoft’s AI Platforms team in Seattle. He has a strong open-source footprint in flagship projects like ONNX and PyTorch, adding LLM-relevant operators (RotaryEmbedding, RMSNormalization) and improving ONNX export for complex autograd scenarios. His background spans embedded vision and mobile ML tooling from internships at Qualcomm to leadership roles in university research and outreach, reflecting both applied engineering and mentorship. A Purdue Computer Engineering graduate (3.75 GPA), he pairs systems-level backend skill with practical ML model interoperability expertise—an undervalued strength that helps bridge research models and production deployments.
8 years of coding experience
Bachelors Degree, Computer Engineering, GPA - 3.75, Bachelors Degree, Computer Engineering, GPA - 3.75 at Purdue University
English, Hindi, Marathi, German, French
Github Skills (16)
neural-network10
pytorch10
machine-learning10
deep-learning10
onnx10
deep-neural-networks10
python10
autograd10
integrations9
exporter9
exports9
tensor9
c-language8
mlops8
cprogramming-language8
Programming languages (3)
Github contributions (5)
onnx/onnx
Aug 2020 - Sep 2021
Open standard for machine learning interoperability
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:46 reviews, 2 commits, 14 PRs in 1 year 1 month
Contributions summary:Shubham primarily contributed to the ONNX repository by implementing new features and fixing issues related to machine learning model interoperability. Their work included resolving segfaults in the ConstantofShape operator and adding a reduction attribute to Scatter style operations. They also added the RotaryEmbedding and RMSNormalization operators, demonstrating their understanding of LLM model components. The user's contributions were crucial for expanding the ONNX standard and improving its support for modern machine learning models.
pytorchmxnetdeep-learninginteroperabilitymachine-learning
pytorch/pytorch
Feb 2021 - Jan 2023
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:ML Engineer
ML EngineerContributions:206 reviews, 34 commits, 82 PRs in 1 year 11 months
Contributions summary:Shubham's contributions primarily focus on enhancing the ONNX (Open Neural Network Exchange) support within the PyTorch framework, specifically concerning the export and optimization of models containing autograd functions. They implemented features like inlining autograd functions and adding support for ATEN_FALLBACK mode, directly impacting the ONNX exporter's ability to handle complex PyTorch models. Furthermore, the user added support for operators like `mse_loss` and `_convolution_mode`, increasing the coverage and compatibility of ONNX export. They also addressed issues regarding scripting and the handling of optional inputs within the ONNX exporter's graph representation.
pythongpu-accelerationdeep-learninggpunumpy
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.