To change one sentence on my site, it used to take me fifteen minutes.
Edit the file, push the change, wait for the deploy, hope nothing broke. All that to change a single word. One day I stopped in the middle of that process and noticed the contradiction: I write that AI changed the way we build, that you can go from the prompt to the product. But my own site depended on a developer workflow to change a comma.
So I built another one. This one you are reading.
The idea: a résumé that is alive
A résumé is a dead document. You write it, freeze it in a PDF, and it is already out of date the moment it is born. I did not want that. I wanted a place where I add an experience, publish a project or fix a text and it is live in seconds, with nobody in the middle.
A "personal WordPress", except mine. Built the way I need it, without depending on anyone else's tool.
And there was one more reason, maybe the main one. I talk about applied AI and building product all the time. If my own site were not proof of that, it would be just another promise. And I have seen enough promises.
The challenges
It is when you open the bonnet that things get interesting.
Build or buy. There are plenty of ready-made tools. I chose to build from scratch, and not out of stubbornness: out of control. When the system is yours, you are not hostage to anyone else's roadmap, and every decision becomes something I can take to the table the next day.
Editing has to be instant. I put a real database behind everything. Change it, and it shows. It sounds like a technical detail, but it is what separates a tool you use from one you avoid. Myself included.
I do not write in code. I insisted on a real visual editor, with a formatting bar and images you drag in, not links you paste. If publishing is tedious, nobody publishes.
The part I am proudest of. The system builds itself. Instead of programming every screen by hand, I describe each section in a single list of fields and the whole panel is built from there. In practice: what would take a day to add now takes minutes. That is what will make this system replicable, and one day, sellable.
And there was one moment that sums up my whole thesis. When it came to rewriting the content, AI did a fast and well-made job. But it wants to make you look impressive: it pulls every number, every result, everything that shines. Part of my job was cutting. What you do not publish, out of discretion or out of respect for whoever trusted you, AI has no way of knowing. It writes. Deciding what stays out is method. And method is human.
What is coming
This is the beginning. The plan is already drawn:
- A media library, so I upload a file once and reuse it everywhere.
- A block-based page editor, to assemble the home page like fitting pieces together, without touching code.
- An editable theme: change colour, font and logo without opening a single file.
- And the part that excites me most: wiring AI into the right places. An analytics panel that does not just show what happened, but suggests what to write. And a studio that takes any piece of content (a post, a project, the evolution of this site itself) and returns the pieces ready for LinkedIn and Instagram.
The trick is that the foundation was built to receive all of that. Every new feature comes in without rebuilding what already exists.
The point
I did not build this site because I am a developer. I am not. I built it because I stopped accepting "it can't be done" as an answer.
Today's tooling lets a marketing manager stand up what used to depend on an entire technology team. So the question has changed. It is no longer whether you can build. It is whether you have the method to use it properly.
I will document every step of this build here. This was the first one.