Smit Hinsu
Software Engineer at Google
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolSmit Hinsu is a software engineer with 12 years of experience focused on machine learning compilers and performance engineering, currently contributing to TensorFlow at Google in Mountain View. He drives backend compiler work across XLA, MLIR and StableHLO, implementing shape inference, legalization passes and bounded-dynamism support that help bridge high-level ML representations to performant HLO. His contributions touch widely used open-source projects (tensorflow, openxla) and include non-obvious infrastructure improvements such as verification tooling and nuanced type/bounds propagation for transpose and reduction ops. At Google he has a strong record of mentorship and cross-team collaboration, earning numerous peer and spot bonuses for impact and helping mentor 20+ engineers. Earlier roles in trading infrastructure and growth teams gave him production-grade experience with low-latency systems, data pipelines and A/B-driven product changes. He combines deep compiler and systems expertise with pragmatic engineering that ships at scale.
12 years of coding experience
4 years of employment as a software developer
B.Tech, Computer Science and Engineering, B.Tech, Computer Science and Engineering at International Institute of Information Technology
Gujarati, English, Hindi
Stackoverflow
11reputation
329reached
1answer
0questions
Github Skills (20)
type-inference10
c-language10
operation10
tensorrt10
machine-learning10
mlr10
intermediate10
multilevel10
deep-learning10
tensorflow10
control-flow10
xla10
compiler10
tensor10
cprogramming-language10
Programming languages (8)
Github contributions (5)
openxla/stablehlo
Oct 2022 - Jan 2023
Backward compatible ML compute opset inspired by HLO/MHLO
Role in this project:
Back-end Developer
Back-end DeveloperContributions:122 reviews, 11 commits, 14 PRs in 3 months
Contributions summary:Smit primarily contributed to the `stablehlo` project by implementing and refining shape functions for various operations, specifically focusing on the `Transpose` and `Reduce` operations. Their work involved modifying the core logic for type inference, especially concerning how bounds are handled and propagated, including updates to the testing framework. They demonstrated expertise in the dialect-specific details within the project, and also worked on the underlying verification infrastructure.
hlomachine-learningcompute
tensorflow/tensorflow
Mar 2018 - Jan 2023
An Open Source Machine Learning Framework for Everyone
Role in this project:ML Engineer
ML EngineerContributions:96 reviews, 704 commits, 9 PRs in 4 years 10 months
Contributions summary:Smit contributed to the TensorFlow repository by implementing and modifying code related to XLA (Accelerated Linear Algebra) and MLIR (Multi-Level Intermediate Representation). Their work involved adding passes for legalizing StableHLO to HLO, introducing and refining segment reduction operations, and fixing issues with dynamic input handling in NMS (Non-Maximum Suppression) legalizations. They also made changes to support bounded dynamism and updated control flow patterns to handle quantized types.
pythondata-sciencedeep-learningmlmachine-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.