Saurabh Chokshi
Engineering Leader at Meta
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolSaurabh Chokshi is an engineering leader with 10+ years building high-performance, scalable full-stack systems across fintech and consumer platforms, currently leading engineering at Meta from New York. His background blends deep systems programming and distributed systems work (including impactful contributions to PyTorch’s distributed checkpointing) with product-focused delivery for payments, credit risk, private equity and wealth management at Goldman Sachs and Facebook. A CFA charterholder who started as a software engineer building low-level video and ASIC interfaces, he uniquely couples quantitative finance domain knowledge with hands-on backend and DevOps expertise. Known for turning ground-up platforms into operationally robust services, he excels at bridging engineering rigor, regulatory-aware design, and measurable business outcomes.
10 years of coding experience
13 years of employment as a software developer
Masters, Computer Science, Masters, Computer Science at University of Southern CaliforniaBachelor of Engineering, Electronics & Communication, Bachelor of Engineering, Electronics & Communication at Hemchandracharya North Gujarat University
English, Hindi
Github Skills (10)
neural-network10
pytorch10
machine-learning10
distributed-systems10
deep-learning10
python10
unit-test10
autograd9
tensor8
gpu8
Programming languages (2)
Github contributions (5)
pytorch/pytorch
Mar 2024 - Mar 2025
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:32 reviews, 8 PRs, 54 comments in 1 year
Contributions summary:Saurabh's commits primarily revolve around enhancements and optimizations within the distributed checkpointing (DCP) system in PyTorch. They introduced new features, including modules metadata and save plan caching, to improve efficiency and reduce collective overhead. Furthermore, the user addressed bug fixes related to all-gather keys and stateful/non-stateful load validation within the DCP framework, directly contributing to the system's reliability and performance. Their work spans from code-level changes to testing enhancements demonstrating a solid understanding of distributed systems and performance optimizations.
pythongpu-accelerationdeep-learninggpunumpy
saumishr/pytorch
Jun 2024 - Apr 2025
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Contributions:58 pushes, 8 branches in 9 months
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.