Tristan Rice
Staff Software Engineer at Meta
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolTristan Rice is a Staff Software Engineer with 14 years building high-performance distributed systems and ML infrastructure, currently working on PyTorch Distributed at Meta in Seattle. He has driven production-grade tooling like TorchX and TorchElastic and maintains key PyTorch distributed components used for large-scale model training. His background spans low-level concurrency and systems work (contributions to Facebook's folly and CockroachDB tooling) as well as developer-facing tooling (Go REPL enhancements), giving him a rare blend of systems rigor and developer ergonomics. Tristan has repeatedly improved reliability and performance in multi-process and GPU-accelerated workloads, and his open-source fixes address subtle issues such as race conditions, interrupt propagation, and packaging edge cases. Collected experience from Meta, Cruise, and internships at Google and others underpins his pragmatic approach to scaling ML infrastructure for real-world, production constraints.
14 years of coding experience
6 years of employment as a software developer
Bachelor of Science (B.Sc.) Computer Science, Bachelor of Science (B.Sc.) Computer Science at The University of British Columbia
English
Stackoverflow
86reputation
6kreached
3answers
2questions
Github Skills (43)
pytorch10
c-language10
python10
futures10
back-end-development10
interpreter10
ast-parser10
distributed-computing10
go10
exception-handling10
parsing10
repr10
parse10
replit10
cprogramming-language10
Programming languages (19)
Github contributions (5)
d4l3k/go-pry
Mar 2015 - Jul 2021
An interactive REPL for Go that allows you to drop into your code at any point.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:179 commits, 56 PRs, 131 pushes in 6 years 5 months
Contributions summary:Tristan primarily worked on improving the functionality of a Go REPL. Their commits show the implementation of an interpreter, improvements to parsing and evaluating expressions, and the addition of features like struct literals and basic math operations. They also implemented key improvements such as a basic form of autocomplete and the ability to add variables to scope, significantly enhancing the REPL's usability and power. The user's work demonstrates a strong understanding of the Go language and its parsing/evaluation capabilities.
golangpointrepldropgo
pytorch/pytorch
Oct 2018 - Jul 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:Back-end Developer
Back-end DeveloperContributions:293 reviews, 57 commits, 114 PRs in 3 years 9 months
Contributions summary:Tristan contributed to the core PyTorch library by addressing several issues related to distributed computing and packaging. Their work includes moving worker registration within the distributed package to avoid duplicate registrations. They also added a workaround to address implicit numpy dependency when calling .numpy() on models loaded via torch.package. The user also fixed issues related to running multi-processing workloads, and implemented various performance optimizations.
pythongpu-accelerationdeep-learninggpunumpy
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.