Mengwei Liu
Staff Software Engineer at Meta
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMengwei Liu is a Staff Software Engineer in Menlo Park with 11 years of experience building high-performance ML infrastructure and production systems. Currently at Meta Superintelligence Labs, he architects real-time inference pipelines, optimizing scheduling, batching, and orchestration for LLMs and vision models. Previously he led efforts to bring PyTorch runtime and interpreter to edge and mobile devices at Facebook, improving modularity and hardware-scalable inference. His open-source contributions to PyTorch’s torchgen—adding custom namespace support, schema optimizations, and Executorch codegen—reflect deep expertise in code generation and runtime tooling for ML frameworks. He has a strong systems and networking background from Oracle and holds an MS in Electrical and Electronics Engineering from UCLA. Colleagues rely on him for pragmatic performance tuning that bridges research models and constrained production environments.
11 years of coding experience
11 years of employment as a software developer
University of California, Los AngelesHong Kong University of Science and Technology (HKUST)
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (8)
management10
gatsbyjs10
code-generation10
manage10
namespaces10
python10
cprogramming-language9
c-language9
Programming languages (5)
Github contributions (5)
pytorch/pytorch
Mar 2021 - Jan 2023
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
Back-end Developer
Back-end DeveloperContributions:293 reviews, 470 commits, 180 PRs in 1 year 10 months
Contributions summary:Mengwei contributed to the PyTorch library, specifically focusing on enhancing the code generation process within the torchgen module. Their work involved adding support for custom namespaces in native function declarations, optimizing schema registration logic, and generating code for Executorch, a separate but related project. The user also improved the unboxing functionality and adjusted the build system, ensuring that the necessary modules and dependencies are correctly linked. These changes likely improve the flexibility and maintainability of the library's internal code generation tooling.
gpu-accelerationneural-networkpythonautogradgpu
larryliu0820/executorch-2
Nov 2023 - Mar 2026
End-to-end solution for enabling on-device AI across mobile and edge devices for PyTorch models
Contributions:89 pushes, 58 branches in 2 years 3 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.