James Wu
Software Engineering Manager at TikTok USDS Joint Venture
California, United States
Summary
🤩
Rockstar🎓
Top SchoolJames Wu is a seasoned software engineering manager with 11 years of experience leading large-scale, reliability-focused teams across AI, search, and fintech. He has driven mission-critical systems at LinkedIn powering 70M+ real-time search requests per hour and led ML-driven safety and risk systems at Meta and TikTok to mitigate bias and harmful content for platforms with over a billion users. His hands-on background spans backend systems, compilers and ML infra—he contributed to PyTorch’s Inductor compiler and improved stability in Facebook’s Flow and HHVM projects—demonstrating deep expertise in low-latency, high-throughput architectures and ML tooling. As a founder/CTO in fintech and blockchain, he built cross-border payment infrastructure and smart contract integrations, blending product-level thinking with platform engineering. Based in California and Berkeley-educated in EECS, he is equally comfortable refactoring GPU kernel code as he is scaling distributed services and mentoring growing teams. A pattern-matcher by trade, he combines production-grade engineering with a pragmatic focus on privacy, compliance, and operational excellence.
11 years of coding experience
22 years of employment as a software developer
B.S, Electrical Engineering and Computer Science, B.S, Electrical Engineering and Computer Science at University of California, Berkeley
Github Skills (31)
pytorch10
caching10
type-check10
python10
technical-writing10
hacking10
machine-learning10
interpreter10
triton10
autograd10
cuda10
compiler10
error-handling10
type-checking10
induction10
Programming languages (10)
Github contributions (5)
facebook/hhvm
Jul 2016 - Sep 2022
A virtual machine for executing programs written in Hack.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:7 reviews, 503 commits, 10 comments in 6 years 3 months
Contributions summary:James primarily contributed to the development of the Hack Virtual Machine (HHVM), specifically focusing on the addition of new language features and refinements. Their work included implementing and improving attributes related to code modularity such as CippGlobal, and later to incorporate the use of readonly declarations. The contributions demonstrate a strong understanding of the codebase and involve modifications to parser, typechecker, and interpreter components of the Hack language. This is evident from the numerous changes related to internal workings, particularly for packages.
virtual-machinehhvmphphacklang
pytorch/pytorch
Dec 2023 - Nov 2025
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:Backend Developer & ML Engineer
Backend Developer & ML EngineerContributions:374 reviews, 229 PRs, 2429 pushes in 1 year 11 months
Contributions summary:James primarily contributed to the development of the Inductor (PyTorch's compiler) project. Their work involved refactoring, adding features and tests related to Triton kernels, CUDA kernels, and Autograd caching. The user focused on various components related to the inductor architecture, including codegen and autograd systems, as evidenced by their contributions.
gpu-accelerationneural-networkpythonautogradgpu
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.