I was quoted €40,000
to build my idea

01 Learning to build

“What the fuck is Laravel?”

Then vibe coding started becoming a thing, and I jumped in almost immediately. Now, here's the funny part: I had absolutely no idea how to code.

I knew Linux. I'd been managing my online business on my own servers for years. I knew enough basic HTML and CSS to avoid asking a friend every time I wanted to change something.

That was pretty much my technical background.

I started small, experimenting with WordPress plugins and building little things to prove to myself that I could actually do this.

And then the journey began.

For the next year, I built almost every day. I spent countless hours working alongside AI agents — often more than 10 hours a day — building, testing, breaking things, fixing them, and starting all over again.

Feature by feature, I began replacing the expensive tools I had been paying for with my own solutions. One thing I discovered along the way became one of my favorite ways to build:

Whenever I wanted to create a new feature, I'd ask my AI agent to search GitHub for open-source projects that already solved a similar problem.

Why reinvent the wheel?

I'll study what's already out there, understand how it works, and adapt the right ideas and solutions to my platform.

And something else I notice happening.

I'd take a break, open social media, and my feed would suddenly be full of research papers, new projects, and people building features remarkably similar to the ones I'd just been working on at the same moment.

Sometimes it felt like my mind and the universe were having a conversation. I feel connected and have fun with synchronicity.

02 AIOHM

Artificial Intelligence for Organic Harmonic Marketing

One year later, the AIOHM Platform was born. Why AIOHM?

Well, my marketing agency is called OHM Events — Organic Harmonic Marketing for Events. So naturally, I gave my platform a name that reflects its purpose:

A little wordplay, perhaps, but I think it fits perfectly. :)

Today, it's been running for months, supporting my own projects and a few client projects. And honestly, the best way I can explain what I've built is to share the journey itself.

Because AIOHM didn't come from a business plan written on a whiteboard, it grew from real problems, real experience, curiosity, and a lot of experimentation.

It's honestly hard to believe that 12 months ago I literally asked an AI: “What the fuck is Laravel?” 😂

If you had told me then that one year later I would be debugging:

01 Graphs

Knowledge graph entity-relationship topologies paired with vector embeddings

02 Gates

Bidirectional human-in-the-loop webhook gates

03 Queues

Multi-tenant database connection lifecycles on native Redis/MySQL queues

…I would have laughed my ass off at it. I wouldn't have even understood the sentence. Building AIOHM transformed how I think about software. AI didn't do the work for me — it acted as the sparring partner that helped me bridge the gap into distributed systems and deterministic agent orchestration.

03 AdiSaan comes back

A name I have signed for thirty years

And that's when AdiSaan came back to life.

AdiSaan — Your online partner.

This name has been part of my identity for almost 30 years. Long before social media, websites, and all the online business stuff, it was the name I used to sign my letters.

Yes, I know. I'm 50 years old. I write letters :) But there's something I've come to appreciate more and more with age:

Experience.

The kind you collect by living

Real experience. The kind you collect by living, working, traveling, making mistakes, meeting people, and seeing the world from different perspectives. And I believe that experience is becoming more valuable in the age of AI.

So AdiSaan has a new mission. First, let me introduce the pieces.

AIOHM is the platform I've been building. It's also home to my autonomous AI agent, Muse. AdiSaan is the public experiment — the place where I want to bring together human imagination and artificial intelligence.

04 #TBWIM

The Best Way I Imagine

And here's the idea behind it:

I want to ask one question every day. Questions about work, money, creativity, relationships, travel, purpose, personal growth, and all the other things that make up this strange and beautiful experience called life.

Then I want to collect one million answers for each question.

Answers from real people. Different experiences. Different opinions. Different ways of imagining what is possible. And AI perspectives too — kept separate, so we can compare what humans imagine with what AI comes up with.

Why am I doing this?

Because I've noticed something over the last year of working so closely with LLMs: AI is becoming incredibly good at producing answers. But the quality of those answers still depends enormously on the questions and ideas we bring to it.

And I worry that we're slowly getting used to outsourcing the imagination itself.

We ask AI what to think, what to create, what to say, and what to do. But what happens when we stop giving it our own original thoughts? What happens when everyone starts asking the same questions and getting increasingly similar answers?

I don't want to fight AI. Quite the opposite.

I want to give AI better human input.

More imagination. More lived experience.

More imagination. More lived experience. More unusual ideas. More perspectives from people who see the world differently.

Better inputs can lead to better outputs.

That's the experiment.

05 Two Muses

Machine Utility for Soulful Expression

And there's one more coincidence I absolutely love. Within the AIOHM Platform, I had already built and named my own AI agent, M.U.S.E. — Machine Utility for Soulful Expression. It's the AI companion at the heart of my platform. It has access to project context and workspace memory, connects to different AI models, helps turn ideas into actionable workflows, and even has a human approval system to keep important decisions under human control.

Then, in September 2026, Meta introduced its own personal AI agent called Muse, initially powered by its Muse Spark model.

I had to smile.

Two independent projects. Same name. Completely different journeys.

So, my M.U.S.E. isn't Meta's Muse, but the coincidence certainly makes the story more interesting. Maybe I should stop trying to explain these coincidences and enjoy watching everything come together. 😂