Tim Zerrell
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolTim Zerrell is a Senior Software Engineer with 11 years of experience building high-performance back-end systems and compilers for AI and zero-knowledge computing. Based in Seattle, he has driven cryptographic and zkVM implementations at RISC Zero and helped transition PlaidML from ad hoc codegen to MLIR-based compiler infrastructure at Intel/Vertex.AI. His work spans low-level math (roots of unity, FRI, Merkle verification) to practical deep-learning ops (transposed convolutions, gradient overrides), bridging research-grade algorithms with production constraints. A mathematician by training (Master’s, University of Washington), he pairs rigorous theory with hands-on Rust and MLIR engineering and maintains personal open-source projects showing curiosities beyond his primary roles.
11 years of coding experience
5 years of employment as a software developer
Master’s Degree, Mathematics, Master’s Degree, Mathematics at University of WashingtonBachelor’s Degree, Mathematics, Bachelor’s Degree, Mathematics at Pomona College
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (20)
convolutional-neural-networks10
python10
machine-learning10
deep-learning10
merkle-tree10
cryptography10
neural-network10
web-framework10
rust10
risc-v9
tensorflow9
keras9
documentation8
algorithm7
algorithms7
Programming languages (6)
Github contributions (5)
plaidml/plaidml
Oct 2017 - Jun 2022
PlaidML is a framework for making deep learning work everywhere.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:289 reviews, 189 commits, 137 PRs in 4 years 8 months
Contributions summary:Tim contributed significantly to the PlaidML project by adding and modifying code related to deep learning frameworks. A key focus was the implementation of transposed convolutions (conv2d_transpose, conv3d_transpose) within the plaidml/keras/backend.py file, involving modifications to core functions and the addition of supporting code for padding. Further contributions involved improvements to EDSL, adding operations like reshape, constant and implementing the gradient overrides, demonstrating a focus on framework development and backend functionality.
pytorchtvmdeep-learningmachine-learningcompiler
risc0/risc0
Jul 2022 - Jan 2023
RISC Zero is a zero-knowledge verifiable general computing platform based on zk-STARKs and the RISC-V microarchitecture.
Role in this project:Back-end Developer
Back-end DeveloperContributions:450 reviews, 11 commits, 187 PRs in 6 months
Contributions summary:Tim primarily worked on core cryptographic and zero-knowledge proof (zkVM) components, including roots of unity, Merkle tree verification, and the FRI protocol. They refactored code to use generic fields and converted Baby Bear field implementations. Additionally, the user addressed documentation compilation warnings and wrote Rust docs for the zkVM, including guides for the guest, build and other related components. Furthermore, they updated the tests and made corresponding code changes.
risc-vcryptographyverifiablestarkszero
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.