AI OS

Build your own AI Operating System to automate your business and life

AI OS shows you how to build the layer that connects AI to your context, your data and your tools. It's platform agnostic, so it works with any agent and runs wherever suits you.

Read chapter 1 first

von Navid Moazzez·Aktualisiert 25. Sept. 2026·5 Min. Lesezeit·

AI OS is a book about building your own AI operating system (AI OS): the layer underneath your AI tools that connects AI to your context, your data and your tools.

It's platform agnostic and adaptable, so it works with any agent and runs wherever suits you.

I run my own business this way. The same system lets me do the work of more than 10 full-time employees by myself, every day.

In this summary, you'll learn:

What the AI OS book is about
The 7 layers of an AI OS
What's inside the book, and the ideas behind it
Who should read it, and my favorite quotes from it
Why I'm writing it
How to read it first

Let's start with what the book is about.

What is the AI OS book about?

Most people use AI as a series of one-off prompts. They open a chat, paste in who they are, explain their offer again and fix the same wrong guesses. Tomorrow they do it all over again.

An AI OS ends that. You tell it once, and every session after that starts already knowing you.

Because it's platform agnostic, it runs wherever suits you: on your computer, on an always-on VPS, in the cloud from a private GitHub repo, or through tools like Grok Bot that make it simpler to run. Its MCP server, CLI and API hook it up to any agent, even cloud ones like Claude Cowork and ChatGPT. The book shows you how to build yours from scratch, one layer at a time.

AI operating system (AI OS): The layer underneath your AI tools that connects AI to your context, your data and your tools. It's platform agnostic: any agent can use it, wherever it runs.

Navid Moazzez

@thenavidm

Die meisten Gründer nutzen KI immer noch wie ein Suchfeld.

Die, die vorne liegen, führen ihr Business mit ihrem eigenen AI OS.

The 7 layers of an AI OS

Every AI OS is built from the same 7 layers. Each layer needs the one below it, so you build them in order, and the book gives each layer its own chapter.

Context

Its job: Knows you
What goes in it: Your rules file and your context files: who you are, what you sell, your voice

Data

Its job: Holds your business in one place
What goes in it: Customers, sales, content and conversations, linked so it can answer questions

Tools

Its job: Reaches your apps
What goes in it: Connectors to your email, calendar, CMS, payments and more

Memory

Its job: Keeps what it learns
What goes in it: Your brain, meetings, decisions, corrections and daily notes

Skills

Its job: Does the work
What goes in it: One folder per job, each with a SKILL.md

Automations

Its job: Runs without you
What goes in it: Daily and weekly routines, and scheduled runs

Apps

Its job: Shows you the answer
What goes in it: A brief page, a dashboard, a command center

Layer	Its job	What goes in it
Context	Knows you	Your rules file and your context files: who you are, what you sell, your voice
Data	Holds your business in one place	Customers, sales, content and conversations, linked so it can answer questions
Tools	Reaches your apps	Connectors to your email, calendar, CMS, payments and more
Memory	Keeps what it learns	Your brain, meetings, decisions, corrections and daily notes
Skills	Does the work	One folder per job, each with a `SKILL.md`
Automations	Runs without you	Daily and weekly routines, and scheduled runs
Apps	Shows you the answer	A brief page, a dashboard, a command center

The test is simple. Open a fresh session and type "write a post about my offer." If it names your offer and your price in your voice, your AI OS is working. If it asks what you sell, start with Context.

What's inside the book

The book gives each layer its own chapter, with a chapter before them and 2 after:

What an AI OS is. Why every AI tool starts from zero without one, and what changes when it doesn't.
Context. Everything your AI OS knows about you before you type a word: who you are, what you sell, who you sell to and how you write.
Data. Your customers, sales, content and conversations in one place, linked so your AI can answer questions about your business.
Tools. Connectors to your email, calendar, CMS and payments, so your AI OS reaches the apps you already use.
Memory. Where your decisions, corrections and daily notes land, so you never give the same correction twice.
Skills. one folder per job, so the work you repeat every week runs from a single command.
Automations. Routines that run without you, like a morning brief that's waiting before you sit down.
Apps. A page that answers before you ask: a brief, a dashboard, a command center.
Your first 14 days. What to expect in your first 2 weeks, from your first setup interview to your first skill of your own.
How to measure your AI OS. One number per layer that tells you whether it's working, and a weekly score out of 100.

The ideas behind the book

Systems over goals

A goal is a wish with a deadline. A system is something that produces the result whether or not you are motivated that day. Everything I build, for me or for other people, is an attempt to move something from the first category into the second.

Context comes first

Every layer after it works better because your AI already knows you.

Keep the tools you use

You don't switch anything to build an AI OS. You connect what you already have.

Build it by hand, then automate it

A skill earns its schedule by working first.

Navid Moazzez

@thenavidm

You will not be replaced by AI. You will be replaced by someone who uses AI. And someone who uses AI will be replaced by someone with their own AI Operating System (AI OS).

Who should read AI OS

Founders, creators and solopreneurs who use AI every day but still start from scratch in every chat.
Anyone who wants AI to know their business: their context, their data and their tools.
People who use Claude Code, Codex or other agents and want one system behind all of them.
Teams of one who want to do the work of a bigger team without hiring one.

Popular quotes from AI OS

Navid Moazzez

@thenavidm

You will not be replaced by AI.

You will be replaced by someone who uses AI.

And someone who uses AI will be replaced by someone with their own AI Operating System (AI OS).

Navid Moazzez

@thenavidm

Your master prompt is your AI identity: who you are, how you write, and what you never say.

Write it once. Use it everywhere.

Navid Moazzez

@thenavidm

The fastest way to get better output from AI is to stop repeating yourself and give it your context once.

Navid Moazzez

@thenavidm

Your first AI setup will be messy. Mine was.

Fix one thing every time it gets something wrong, and it gets better every week.

Why I'm writing it

I'm the founder of AI OS Lab, and the AI OS is the through line in everything I do now. It's what I write about every week, and what I build with people directly.

Your first setup won't be perfect. Mine isn't either. The more you run it, the more it knows, and the book is how I've learned to build it, one layer at a time.

How to read it first

Everyone on the list at aiosbook.com reads the first chapter before the book goes on sale.

Learn more

How to build an AI OS: My step-by-step guide to building your AI OS in Claude Code, Codex or any AI agent, from an empty folder to a system that runs every day.
The AI OS book: The book's own site, where you join the list and read the first chapter first.
How I use AI: The tools, agents and workflows I run my own business on, every day.
About Navid: My story, from law school in Stockholm to building my business on AI.

FAQs about the AI OS book

Have questions about the AI OS book? Here are answers to the most common ones.

What is the AI OS book about?

It shows you how to build your own AI operating system (AI OS), the layer that connects AI to your context, your data and your tools, one layer at a time. It's platform agnostic, so it works with any agent and runs wherever suits you.

What is an AI operating system (AI OS)?

An AI OS is the layer underneath your AI tools. It holds your context, your voice, your offers, your customers and your workflows in one structured place that every AI you use can read.

Is an AI OS an actual operating system?

No. It runs on top of the computer you already have, inside an AI agent like Claude Code. The name comes from what it does: the way macOS loads your files and settings when you log in, an AI OS loads your business every time a session starts.

What are the 7 layers of an AI OS?

Context, Data, Tools, Memory, Skills, Automations and Apps. Each layer needs the one below it, so you build them in order.

Do I need to code to build an AI OS?

No. An AI OS is plain text files and a few commands. You write the files in plain English, and the agent handles any technical setup for you.

Which AI agents does it work with?

Any agent that reads files from a folder. Claude Code reads CLAUDE.md, Codex and Cursor read AGENTS.md, and Gemini CLI reads GEMINI.md. Make those the same file and they all share one AI OS.

Does an AI OS only run on my computer?

No. It can run on your computer, on an always-on VPS, in the cloud from a private GitHub repo, or through tools like Grok Bot that make it simpler to run. Its MCP server, CLI and API also connect it to cloud agents like Claude Cowork and ChatGPT.

Is an AI OS the same as a second brain?

No. A second brain stores what you know. An AI OS puts it to work. Many setups point the AI OS to a second brain folder, so the agent can search your notes when a task needs them.

What is the difference between an AI OS and an AI agent?

The agent is the tool that does the work, such as Claude Code. The AI OS is what the agent reads: your context, your connectors and your commands. Switch agents and your AI OS comes with you.

Where does my data live?

Wherever you run it: in plain files on your computer, on an always-on VPS or in a private GitHub repo. API keys stay in a settings file that git ignores. What gets sent to the AI model depends on the agent you use, so check its data settings.

What does an AI OS cost?

The AI OS files are free. You pay for the AI agent you run it in, such as a Claude or ChatGPT plan, and for any paid tools you connect.

Can a team share one AI OS?

Yes. Keep the shared context and commands in one folder the team syncs, such as a private GitHub repo, and let each person add their own files on top.

Can I read the book before it comes out?

Yes. Everyone on the list at aiosbook.com reads the first chapter before the book goes on sale.