• Led a study group for Software Foundations, a book series on automated proofs for software.
• Collaborated on software in Rust, C/C++, and JavaScript: an LLVM analysis pass, an in-browser text-based adventure game, voice control software, and more.

• Team lead for TPUv7, the successor to Trillium.
• Delivered core features for high-performance machine learning: embedded firmware and Linux drivers for libtpu, in C/C++.
• Delivered software to accelerate chip delivery: behavioral models of hardware, code generation.
• Led engagements with other teams:
  • Hardware: hardware/software co-design, behavioral specification of hardware.
  • Video accelerator software (VCU): shared platforms for software development.
  • Compiler (XLA): low-level features to support machine-learning runtimes.

• Improved security for the platform: designed identity architecture for Anthos Agent, using workload identity to protect key confidentiality.

• Supported systems for customer-facing load-balancing and software-defined networking.
• Provided Tier 1 oncall (<2 min response), incident management, and postmortem ownership.
• Maintained and presented SRE EDU curriculum for networking.

Provided tutorial, grading, and material updates for Program and Data Representation.

Community Contributions

Artifact Evaluation Committee Member – IEEE/ACM MICRO 2023 Artifact Evaluation Committee Member – ACM ASPLOS 2022 Evaluated reproducibility of submitted artifacts. Volunteer position.

Author of amusements – SIGBOVIK (satirical conference) 2020 and 2024.

Education

B.S. Computer Engineering – University of Virginia, 2014

CS240 Operating Systems – Stanford Center for Professional Development, 2020