The Bank, the XML, and the First Spark
The best ideas don't arrive at conferences. They arrive in the waiting.
Realschule
Let me say something clearly, before we go any further.
I never finished university. I left mechanical engineering when the math became obvious. My highest formal qualification is Realschule – the middle track of the German school system. Not Gymnasium. Not university. Not an academic title of any kind.
I want to say this clearly because what comes next matters more if you know this first.
At forty-something years old, as a Realschule graduate with no academic degree, I was hired as a freelancer at one of Germany's largest banks.
I managed projects. I trained staff. I optimized systems that handled billions.
Nobody asked to see my diploma. Because the work spoke.
If you are reading this and thinking you are not qualified enough – I want you to sit with that for a moment.
Qualified for what, exactly? Says who?
The Path to Frankfurt
The road to Dresdner Bank was not direct.
Before the bank there was DuPont de Nemours – my first serious freelance project. Then Debitel in Stuttgart – a Mercedes-Benz subsidiary – where I was contracting for Sybase , where I somehow ended up responsible for over 500 customers.
Each project taught me something. Each project made the next one possible.
Not because of certificates or degrees. Because of work. Because of showing up. Because of doing the thing until you understood it better than anyone in the room.
A headhunter found me for Dresdner Bank.
Headhunters don't care about your diploma. They care about your track record.
Mine was enough.
The Waiting
My job at the bank was to maintain an application.
Maintain. Not build. Not design. Maintain.
The application involved a query that touched fifteen database tables. Fifteen. A single SQL statement of magnificent complexity, built over years by people who were probably no longer there.
I optimized it stacking views and stored procedures. Not because anyone asked me to. Because it was inefficient and inefficiency bothers me. And because I had time.
That is the part nobody tells you about large corporate projects:
There is a lot of waiting.
Systems need to run overnight. Deployments need sign-off from seven departments. Meetings get scheduled to discuss the meetings. Approvals travel through hierarchies like messages in bottles.
And in that waiting – in those long Frankfurt afternoons when the application was running and the reports were generating and there was genuinely, absolutely nothing to do –
I opened a new project.
Swing and XML
The year was somewhere in the early 2000s.
Java Swing was how you built desktop applications. XML was the language everyone was excited about – structured, readable, a way to describe things without hardcoding them.
I had an idea.
What if you could describe a user interface in XML – declare its structure, its components, its layout – and then render it at runtime?
Not write imperative code for every button and every panel. Just describe what you want. And let the system figure out how to show it.
I started building it.
In the waiting. Between the queries. During the overnight runs. Nobody asked me to. Nobody knew I was doing it. I was just following the pull.
The same pull as Deluxe Paint. The same pull as the Sprite Editor.
Here. This is it. Pay attention.
Open Source
At some point I realized something.
This tool I was building – this XML-driven UI renderer – it worked. Not perfectly. Not completely. But it worked. And it solved a problem that wasn't just mine.
Other developers were writing the same boilerplate. Other developers were hardcoding the same layouts. Other developers were sitting in their own waiting rooms, solving the same problems from scratch.
So I made it Open Source.
Not for recognition. Not for money. Because the thought arrived, simple and clear:
Other people could use this too.
That thought – that small, quiet, obvious thought – was the first Forge thought.
Not the technology. Not the architecture. Not the name.
Just: what I am building could help someone I will never meet.
That is where it started. In a bank in Frankfurt. In the waiting.
Leszek and Leonid
Two years at Dresdner Bank.
In the second year, things changed. I was no longer alone, no longer waiting. I had a partial project leadership role. And I had people to train.
The first was Leszek.
Leszek had an academic title in mathematics. A proper one. The kind that takes years and dissertations and committees. He had been at the bank longer than I had.
I trained him.
A Realschule graduate training a mathematician at a major German bank.
I want to be careful here – Leszek was brilliant. His mathematical mind was extraordinary. But brilliance in one domain does not automatically transfer to another. He needed someone to show him the practical path. The short path. The path that skips the theory and goes directly to what works.
That was apparently me.
The second was Leonid – a young student, Wirtschaftsinformatik, just starting out. Eager. Thorough. The kind of student who, when you point him at a topic, reads the entire textbook.
Every word. Every footnote. Every appendix.
I watched him do this and finally said:
"Leonid. You don't need to read the whole book. Just pick out the important parts. The Rosinen."
He looked at me like I had suggested skipping church.
But I meant it. I have never read a complete book in my life. I read the chapters I need. I find the idea I'm looking for. I move on.
This is not laziness. This is precision.
You don't need everything. You need the right thing.
Leonid eventually understood. Leszek eventually understood.
The Realschule graduate had taught the academics something the university never covered:
Pragmatism is a skill. And it might be the most important one.
What the Bank Actually Gave Me
I left Dresdner Bank after two years with several things:
SQL skills – real ones, earned by optimizing a fifteen-table query out of pure boredom.
Project experience – leading people, training people, navigating the beautiful absurdity of large organizational structures.
The prototype – an XML-driven UI renderer, built in the waiting, now Open Source, now in the world.
The thought – other people could use this too.
But most importantly, the bank gave me something I didn't recognize as a gift until much later:
Two years of proof that the idea worked.
Not perfectly. Not as Forge. Not as anything that would have impressed anyone at a tech conference.
But the core was there. The principle was there.
Describe the structure. Let the runtime handle the rest.
I just didn't know yet that I would spend the next twenty years returning to that idea. Refining it. Failing with it. Walking away from it. Coming back to it.
Until one February morning in 2026, when I sat down and said:
Third attempt. This time we finish it.
The Pattern, Again
Dupont. Cybase. Debitel. Dresdner Bank.
Each one a waiting room. Each one a classroom. Each one a building site in disguise.
The corporate world thinks it is buying your time.
It is actually funding your education.
And occasionally – if you are paying attention – it is funding your next project.
I was paying attention.
Lektion: Die beste Zeit zum Bauen ist die Zeit die dir jemand anderes bezahlt. Nutze die Wartezeit.
Next: Chapter 4 – Too Much, Too Soon
