Jack Montgomery
Senior Staff Software Engineer at Meta
Los Altos, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJack Montgomery is a Senior Staff Software Engineer with 13 years of experience, based in Los Altos, California, currently driving Meta’s AI acceleration efforts and ads delivery infrastructure. He combines deep systems and compiler expertise with practical ML engineering, having contributed significant backend work to the widely used PyTorch/Glow compiler—optimizing tensor ops, adding quantized convolution support, and improving memory efficiency and ONNX compatibility. Jack excels at bridging research-grade ML models and production hardware runtimes, turning complex numerical and memory-management challenges into robust, test-covered features. His profile reflects a focus on low-level performance and correctness at scale, with a knack for making cutting-edge ML stacks interoperable with industry standards.
13 years of coding experience
Carnegie Mellon University
Stackoverflow
2,773reputation
246kreached
0answers
20questions
Github Skills (27)
c-language10
operation10
tensorrt10
onnx10
tensorflow10
optmization10
optimisation10
compile10
neural-network10
network10
tensor10
cprogramming-language10
data-structure9
algorithm9
algorithms9
Programming languages (7)
Github contributions (5)
pytorch/glow
Nov 2018 - Mar 2022
Compiler for Neural Network hardware accelerators
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:5 reviews, 276 commits, 310 PRs in 3 years 4 months
Contributions summary:Jack primarily worked on refactoring and extending the Glow library related to tensor operations, specifically focusing on optimizing and adding support for various operations within the context of neural network compilation. This included modifying and extending the functionality of existing tensors, such as adding the Modulo operation, and improving memory management for efficiency. Furthermore, the user added a suite of tests for accuracy checks, along with ONNX model writing and loading for quantized operations, making Glow compatible with ONNX models that use quantized convolutions.
compilerhardwareneural-network
jackm321/glow
Nov 2018 - May 2021
Compiler for Neural Network hardware accelerators
Contributions:685 pushes, 257 branches in 2 years 7 months
hardware-acceleratorscompilerneural-networkacceleratorshardware
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.