shunzz.com
Spec-driven × AI agents × full-stack delivery. I don't just use AI to write code — I manage how AI works, and I build tools for it.
01
Japanese Literature major turned AI-Native Software Engineer (JLPT N1). I ship enterprise systems, ML models and cloud infrastructure with a spec-driven, AI-agent workflow; I write Claude Skills, built an MCP Gateway, and teach the workflow internally. In the AI era an engineer's value is judgment and process design, not typing speed.
Japanese Literature degree, JLPT N1. My career began with language; a six-month AI/cloud bootcamp moved me into engineering. I built AI applications at a startup (content-to-video, LINE avatars and voice cloning) and now work as a Senior Systems Analyst at a retail group's IT subsidiary, spanning data engineering, DevOps and cloud architecture.
Over the past year I moved my workflow to spec-driven development with AI agents: I adopted the open-source superpowers workflow (brainstorm → design spec → implementation plan → TDD), where AI executes and I review and decide, then tuned it for real projects — one shared AGENTS.md driving two agents, an automated AI code-review loop, and project-specific skills added to the flow (e.g. an Azure DevOps PR skill). With it I independently delivered an enterprise allocation platform (2.5 months), the AI service of a Customer Data Platform, demand-forecasting ML models, a company MCP Gateway in 3 days, and a full app with 62 tests in one day. Every number is in the git history.
I don't just use AI tools — I package team domain knowledge as Claude Skills with evals, built the gateway that lets every AI tool in the company reach internal APIs safely, and teach the workflow in internal talks. Next: build something that actually matters.
02
03
AI job-search assistant: upload résumé → AI diagnosis → auto-crawl listings → fit ranking → tailored cover letters
Threads AI roast generator: enter a handle → four roast styles as share cards → share to unlock hidden styles
Product thinking and a lesson in failure: Threads' anti-scraping returned 200 with gutted content, so the AI fabricated personalities. "Never fabricate" became a product rule. It spread less than Career Agent — tools that solve real problems travel further than entertainment.
Wraps internal enterprise APIs as MCP tools so any AI coding agent reaches them safely through one entry point
Design focus: per-client auth, SSRF protection, encrypted upstream credentials, audit logs with sensitive-parameter masking. Tool catalog lives in a DB — approved tools go live without code changes or restarts.
Retail allocation platform, CDP AI service, demand-forecasting ML, Azure IaC
Delivered solo with AI collaboration at a retail group's IT subsidiary: a 2.5-month allocation platform (FastAPI + React 19), a CDP AI service on a Google ADK pipeline, LightGBM/CatBoost allocation models, and modular Bicep infrastructure.
04