Skye Wanderman-milne
SWE at Google
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolSkye Wanderman-milne is a seasoned software engineer with 13 years of experience building high-performance back-end systems, currently on the JAX team at Google after earlier work on TensorFlow. They specialize in ML compilers and numerical libraries, contributing substantial low-level improvements such as FFT primitives, PJRT C API enhancements, and memory allocator integrations across flagship projects like XLA, JAX, and TensorFlow. Skye’s code has helped bridge CPU/GPU/device interactions and improved batching, threading, and memory layouts—work that quietly enables faster and more reliable ML workloads at scale. An MIT-trained engineer, they blend rigorous academic foundations with practical production impact, having moved between core infrastructure and research-adjacent tooling for over a decade. Based in San Francisco, Skye is an active open-source contributor whose changes often touch subtle but critical aspects of ML runtime performance and correctness.
13 years of coding experience
1 year of employment as a software developer
BS, Computer Science, BS, Computer Science at Massachusetts Institute of Technology
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (25)
dft10
c-language10
compilation10
python10
memory-management10
machine-learning10
numpy10
fft10
tensorflow10
compile10
xla10
jax10
cprogramming-language10
linear-algebra10
apidoc9
Programming languages (11)
Github contributions (5)
jax-ml/jax
Apr 2020 - Jan 2023
Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
Role in this project:
Backend Developer
Backend DeveloperContributions:40 releases, 527 reviews, 186 commits in 2 years 9 months
Contributions summary:Skye primarily focused on enhancing the JAX library by addressing bugs and extending its functionalities. Their contributions included fixing a bug in convert_element_type for complex dtypes, refactoring existing code into new files to improve maintainability (e.g., extracting parallelization functionality), and integrating support for new primitives, such as the initial FFT implementation. These changes improved the stability and functionality of the library by adding new support for numerical and mathematical functions.
pytorchpythonjitautomatic-differentiationgpu
openxla/xla
Apr 2019 - Jan 2023
A machine learning compiler for GPUs, CPUs, and ML accelerators
Role in this project:Back-end Developer
Back-end DeveloperContributions:9 reviews, 98 commits, 3 PRs in 3 years 9 months
Contributions summary:Skye primarily contributed to the XLA Python client, enhancing its functionality with features like FFT support, BFCAllocator implementation, and device-related improvements. They added extra failure information in cases where some but not all replicas fail. The user's work also involved creating intra-op thread pools for CPU operations such as FFTs, alongside modifications to the compilation process and device handling within the client. The changes also addressed memory allocation with environment variables to optionally use the BFCAllocator.
compilercommunity-drivenmachine-learningmodular
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.