001 — INTRO

1-min intro

006 — WHO

Who I am

Candid

I don't do flattery. I'll tell you the real state of a project, what problems I'm facing, and my actual judgment — even when the answer isn't pretty.

Genuine

No persona, no hustle-bro act. Just a regular person trying to make things work — honest with friends and collaborators alike.

Humorous

Work is serious; people don't have to be. I like finding the funny angle in serious things to make the process less brutal.

004 — BACKGROUND

Where I come from

11 years of combined product and technical experience. From automotive maps to smart cockpits to AI/LLM, I've been making product decisions at the front line of each tech wave.

In December 2025, I voluntarily left my job — not because I was laid off, but because I believe AI will rewrite the rules of every industry in the next 10 years. I wanted to be building before the rules get rewritten.

Now I'm building a one-person company operating system (OPC OS) from scratch using AI. This isn't a side project — it's my full-commitment bet on what the future of work looks like.

002 — NOW

What I'm building

6projects
10agents
110skills

Active projects

005 — PITFALLS

Mistakes I have made

Treating tools as the goal

BACKGROUND
I spent six months chasing every new AI framework, plugging in the latest tool, and barely shipping a single working business flow.
LESSON
Tools are means, not ends. Find a real customer problem first, pick the simplest tool that solves it, then optimize.
NOW
Before adopting anything new I ask: which concrete blocker does this remove? No answer means no adoption.

Building demos, not a business

BACKGROUND
I shipped a stream of flashy demos that earned likes, but not a single one earned a paying customer. Income ¥200/month, burn ¥7,800.
LESSON
Cool does not equal willing-to-pay. Find the buyer first, then build the product. Never the other way around.
NOW
Every project must answer: who pays, why, and how much? No answer means it is a hobby project, not a launch.

Trying to do everything at once

BACKGROUND
I started six projects in parallel. Three months later all six were half-finished and zero were live.
LESSON
Focus beats effort. Do less, finish one thing, then start the next. Half-baked products have zero value.
NOW
One main project per day; everything else goes to backlog. Ship one before starting the next.

S6.5 — Notes

Leave a note for Ian

Log in to leave a note.

005 — INFLUENCES

Who I follow

007 — CONTACT

Get in touch

If you're interested in AI startups, one-person companies, or product exploration — or want to talk about collaboration — I'd love to hear from you. I take every message seriously.