m-naoki-m

About

I work full-time as a software engineer while building personal products on the side —
as a way to keep up with new tech and learn by shipping.
The way I work is built around AI, end to end from design through operations.

See projects Read writing

Profile

Agents handle generation and execution. I set the spec, verify the result, and decide what to publish.
iOS apps, web services, and products that run on AI — from concept to operations.
Start small, validate, and keep refining what proves valuable.

9
Projects
iOS / Web
Domains
19+
yrs of development
AI-native
Workflow

How AI-native development works here

AI-native means building the development process around AI. Agents handle generation and execution; I set the spec, verify the result, and decide what to publish.

Write the spec
Pin down the question, the constraints, and the definition of done.
Let it generate
Agents produce the implementation, the tests, and the docs.
Verify
Since I didn't write it, I check the behaviour instead of assuming the intent.
Fold it back
Trace the cause back to the spec, then generate again.
Raise autonomy
Expand automation where it has proven reliable, so only exceptions reach a person.
I expand autonomy step by step, but I always decide what to publish and carry the final responsibility. The work is designing systems that surface deviations, not reviewing every change.

Domains

Three domains I work across — designing, building, and operating products.

iOS

Designing and shipping iOS apps with SwiftUI / Combine / SwiftData

Web

Building web services and dashboards with Next.js / TypeScript / NoSQL

LLM

Products that run on AI — model selection, prompts, and guardrails

Strengths

  • Designing for real production, not stopping at PoCs
  • Owning end to end from design through operations
  • Working across AWS, generative AI, iOS, and the web
  • Building the evaluation set before widening what runs automatically
  • Designing systems so deviations surface, rather than reviewing every change
  • Project management at multi-million-yen scale, with vendor coordination
  • Hands-on experience running personal blogs and YouTube — SEO to monetization
  • Implementation across control systems, web, SaaS, and generative AI
  • Picking up new tech quickly and applying it in practice

Career

2007
Built control systems in C / Linux; led teams up to 9
2014
Worked across video streaming, internal systems, and virtualization (web / iOS / infra)
2017
Launched personal blogs and YouTube; grew SEO traffic to 30K monthly views
2019
Led a JPY 300M data-platform project as PM
2022
Led a SaaS launch as the de-facto lead engineer
2023
Tech lead for AWS + generative AI systems built for real production
2024
Released Chrono Tap, an iOS app, as a side project
2026
Moved the article pipeline to full automation; review now runs through two models that must agree
Now
Holding all current AWS certifications; back to personal product engineering

Principles

  • Move fast with AI; turn it into value through human judgment
  • Take responsibility through operations, not just the release
  • No hype — be clear about what's possible and what isn't
  • Separate decisions that can be automated from ones that need a human
  • Decide up front what AI does not touch
  • Build small, refine through real use
  • Take user problems seriously
  • Stay transparent and improve continuously

External links

Where I share my work and learnings.

X

Daily notes and engineering thoughts

GitHub

Source code and experiments

Qiita

Technical notes from practice

Zenn

Engineering articles and learnings