const saigetsu = {
role: "AI Product Engineer",
quests: ["workflow automation", "developer tools", "small product experiments"],
superpower: "turning ambiguity into shippable systems",
fuel: ["coffee", "curiosity", "fresh powder"],
rule: "keep humans in control of consequential AI actions",
status: "building in public 🚀",
};我喜欢把模糊的业务流程,炼成可运行、可部署、可验证的软件。
曖昧な業務フローを、実行・検証・改善できるプロダクトへ。
|
Turns fuzzy requests into deterministic, reviewable workflows—with AI helping and humans holding the launch key. |
A six-role delivery harness for AI-assisted Japanese SI projects, complete with review gates and guardrails. |
|
Browser-ready freight calculators and reusable formulas for dimensional weight, pallet loads, and LCL W/M. |
|
🧪 Proof of work
| Project | What is implemented | Verification |
|---|---|---|
| AI Workflow Automation | React, FastAPI, workflow engine, GitHub/OpenAI integration, human checkpoints | Production frontend build and backend health smoke test |
| Harness JP SI | Six roles, multi-platform bootstrap, hooks, skills, templates, governance rules | 75 structural checks passing |
| ShipMathLab | 20 live tools, guides, SEO pages, public formulas, embeddable calculators | Formula tests passing; live public deployment |
I am interested in AI product engineering, workflow automation, developer tooling, and full-stack roles where product judgment matters alongside implementation.
For project-specific questions, open an issue in the relevant repository. Pull requests, curious questions, and slightly over-engineered ideas are all welcome. 🛸
🏂 Carving through messy workflows, one clean commit at a time.


