Pete Walsh is a machine learning engineer with 11 years of experience building end-to-end NLP systems and productionizing ML at scale, currently a Member of Technical Staff at Microsoft AI after leading research engineering at AI2. He combines strong research rigor with pragmatic engineeringâshipping production tokenizers and CRF improvements in top open-source projects like Hugging Face tokenizers and AllenNLP while also contributing core Rust backend work (e.g., rusty-celery). Comfortable across Python and Rust, he bridges model development, build/release automation, and tooling integration, and has a knack for reproducible research and developer ergonomics (Neovim/Obsidian integrations). Based in Portland, he pairs advanced math/statistics training with hands-on open-source impact, and quietly emphasizes maintainability and testing in areas that often get overlooked in ML systems.
11 years of coding experience
6 years of employment as a software developer
Graduate Coursework, Mathematics and Statistics, Graduate Coursework, Mathematics and Statistics at University of Washington
Candidate for Doctor of Philosophy (Ph.D.), Co-major: Mathematics and Statistics, Candidate for Doctor of Philosophy (Ph.D.), Co-major: Mathematics and Statistics at Iowa State University
Bachelor of Science (B.S.), Double Major: Mathematics and Economics; Minor: Statistics, Bachelor of Science (B.S.), Double Major: Mathematics and Economics; Minor: Statistics at St. Lawrence University
Semester Abroad, Mandarin Chinese, Economics, Semester Abroad, Mandarin Chinese, Economics at East China Normal University
đŠ Rust implementation of Celery for producing and consuming background tasks
Role in this project:
Back-end Developer
Contributions:13 releases, 75 reviews, 321 commits in 3 years
Contributions summary:Pete primarily contributed to the back-end functionality of the `rusty-celery` project, which implements a Celery task queue in Rust. Their work involved adding structure, core application features, updating the API, supporting prefetch counts, and refactoring existing code. The user's code changes included the creation of essential components and modifications to the task management system of the project.
Contributions:221 reviews, 136 commits, 339 PRs in 5 months
Contributions summary:Pete implemented core features for the Obsidian.nvim plugin, focusing on the integration of Neovim with Obsidian. They added new commands for interacting with Obsidian vaults, particularly the `:ObsidianOpen`, `:ObsidianNew`, and `:ObsidianBacklinks` commands. The work included creating the necessary infrastructure for the plugin, including its build and testing setup, implementing features such as auto-completion and frontmatter handling, and refining UI integration. Overall, the user contributed significantly to the plugin's core functionality.
luaneovim-luanvim-cmpobsidian-mdneovim
Find and Hire Top DevelopersWeâve 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.
Request Free Trial
Pete Walsh - Member Of Technical Staff at Microsoft AI