I bridge theory and practice — applying computer science theory and algorithm analysis to real-world systems, from chip architectures to cloud infrastructure.
I'm interested in AI accelerators, low-level programming, algorithms, and AWS infrastructure.
- 2025 Japan AWS Jr. Champions
- 2026 AWS Community Builders (Machine Learning)
- AtCoder Blue rated competitive programmer
| Date | Title | Topic |
|---|---|---|
| 2026-02 | Neptune Analytics SSSP Δ-parameter | Delta-Stepping parameter tuning grounded in Meyer & Sanders (1998) |
| 2026-02 | Introduction to RAG | RAG architecture and bottleneck analysis |
| 2026-01 | Do State Machines Dream of Turing? | Turing-completeness of AWS Step Functions via 2-counter machine construction |
| 2025-12 | DeepRacer Cup Finals — Breaking 30 Seconds | 2nd place, reinforcement learning strategy (Jr. Champions) |
| 2025-12 | Trn3 UltraServer | AWS Neuron chip topology evolution for MoE workloads |
| 2025-12 | kani | Formal verification of Rust standard library safety |
| 2025-11 | Fine-tuning Hands-on | SageMaker fine-tuning workshop (Jr. Champions) |
| 2025-11 | Understanding AI Agents and Choosing Dev Tools | AI agent concepts and development tool selection |
- math-optim-rs — Math & optimization algorithms in Rust, verified with Library Checker
- theory-of-computation-with-aws-step-functions — Code for the Step Functions Turing-completeness talk
- nix-config — Reproducible development environment with Nix flakes and home-manager