Build a Software Platform With No Budget:11-Month Rebuild

You can build a software platform with no budget: I lost €250,000 in Saudi Arabia — fifteen years of savings, gone in months — and then built a platform that would have cost €2–3 million the traditional way, alone, in eleven months, for €3,250 in tools. The loss and the build are the same story. You can't understand one without the other, so here is the whole arc, including the parts that don't flatter me.

Fifteen years of doing everything right

I left Portugal at twenty-six with everything I owned in one bag and €5,000 in my pocket. In Helsinki, where the average software engineer made around €3,000 net a month, I was offered €1,300. I took it — desperation doesn't negotiate. Then Germany, better titles, better money, the same corporate politics wearing a different suit. In 2015 I finally got a boss who was actually good; together we turned a struggling operation around, and having proven we could fix it, I got bored of the ceiling above it. I quit and started Evolution4all, my consultancy. Great clients, good money, five published books, talks on four continents.

In 2020, a consulting engagement took me to STC Pay, a Saudi digital bank, for six months. What I found electrified me: an entire ecosystem being built from scratch, and my frameworks translating perfectly into a completely new market. COVID forced me back to Portugal, but Saudi Arabia stayed in my mind for three years — not as a place I'd been, but as a place I'd return to.

Build a software platform with no budget: the October 2023 bet

I went back carrying everything I'd built: a quarter-million euros in savings, my books, my frameworks, fifteen years of hard-earned reputation. The plan was simple — land consulting contracts, do what I'd always done, in a market I'd already validated.

The reality: in the three years since my first visit, the ecosystem had completely pivoted. Vision 2030 had transformed the landscape. Nobody wanted digital transformation consultants anymore — the market wanted startup mentors, accelerator programs, venture capital infrastructure. My entire plan was wrong for this market, and I'd already committed everything to it.

The slow-motion catastrophe

The €250,000 didn't disappear in a dramatic crash. It bled out over months: business setup costs, living expenses, marketing campaigns that generated nothing, pivots that went nowhere. That's the version of failure nobody warns you about — the one slow enough that on any given week, you can still tell yourself it's about to turn around.

I reached out to hundreds of people from my fifteen-year career. Colleagues I'd helped. Clients I'd transformed. Conference "friends" with my business card in a drawer. Out of hundreds, a handful actually helped. The rest went silent or offered encouragement that cost them nothing. Fifteen years of relationship-building, and when I actually needed those relationships, they evaporated like fog.

I was forty-five, in a foreign country, with no savings, no network that could help, and no clear path forward. Sleepless nights. Mornings without a reason to get up. And the voice on repeat: "You're forty-five. You failed. It's over. Go back to Europe. Get a job." It was persuasive because it was partly right. I had failed, by any reasonable metric. The question was whether failure at forty-five meant the story was over.

Then my mother did something that changed the trajectory — she bought me the Claude subscription I couldn't afford. That story has its own post, and I won't cheapen it by summarizing. What matters here is what came after.

The bookThe Solopreneur RevolutionLuis Gonçalves on building a solopreneur business with AI while employed.Get the book →

How I built a software platform with no budget: nothing left to lose

The idea had lived in my head for years: startups drowning in fifteen-plus fragmented tools, accelerators limited by geography, investors trapped in narrow networks — what if all three connected on one platform? The difference now was that I had nothing left to lose, a tool that could multiply my output by an order of magnitude, and a debt to my mother I intended to repay.

Eleven months at a kitchen table, fourteen to sixteen hours a day. In month two I made a decision that felt insane: design the entire system upfront — 964 database tables mapped before a single line of application code. Two weeks of pure design that felt agonizingly slow and saved me months of refactoring. The Product Requirement Prompt methodology was born from the same necessity: I couldn't afford mistakes, couldn't waste time refactoring, had no one to catch my errors. Every feature had to work the first time.

By the end: 964 database tables, 633 service files, over three million lines of code, ten integrated products. Total tool cost: €3,250 — the exact tool-by-tool breakdown lives here, so I won't retell it. Industry estimates for the traditional version: ten to fifteen developers, two to four years, €2–3 million minimum.

What losing €250K taught me that winning never did

Constraints breed creativity in ways that resources never do. I had no money, no team, no co-founder, no runway. Those constraints forced me to invent the PRP methodology, the architecture-first approach, the validation loops that caught errors before they compounded. If I'd had €100,000, I would have hired developers and never discovered any of it. Desperation is an underrated teacher.

Your network is weaker than you think. Test it before you need it. The relationships that survive are the ones built on delivered value, not exchanged business cards — and there are far fewer of them than your LinkedIn suggests.

Failure is expensive, but temporary. Regret is permanent. If FIKR Space had failed — eleven months for nothing usable — I'd have lost eleven months. Painful. But going back to corporate and spending twenty years wondering what might have been? That wondering follows you to your grave. I'd rather fail trying than succeed at something I don't care about.

You don't need the catastrophe

People ask if it was worth it, and the honest answer depends on when you catch me. On the dark days — permanent-feeling isolation, a back wrecked by a kitchen chair, eyes burning from sixteen hours of screen light — no. On the days I watch the platform run and think about the alternative of twenty more years in meetings where nothing gets decided — worth every sleepless night.

But here's the part that matters for you: I built from desperation because I'd burned my safety net first. You don't have to. The same €3,250 economics work from a salary, on one to two hours a day, with your savings intact and your downside near zero. My sequence — bet everything, lose it, rebuild from rubble — is the origin story. It is not the method. The method is the opposite: validate first, build while employed, and let the numbers tell you when to leave.

12 weeksYour corporate career was never the planApply to the Solopreneur Accelerator.Apply now →

Frequently asked questions

How do you lose €250,000 without noticing?

Slowly. It wasn't one bad decision — it was setup costs, living expenses, marketing that generated nothing, and pivots that went nowhere, month after month, in a market that had pivoted away from what I was selling. Slow failure is more dangerous than fast failure because every week feels like the one before the turnaround.

Why didn't you just go back and get a job?

The voice suggested exactly that, nightly. Three things stopped me: I'd already survived worse than a bad week; I had months of shipped progress proving the methodology worked; and the problem I wanted to solve was still real. When I nearly quit over a broken feature in early 2025, I made the problem small — fixed one function, then the next — and by the end of the weekend the feature worked.

How is a €3,250 build cost even possible?

AI executes; you direct. I designed the architecture, wrote detailed specifications, reviewed every line, and Claude generated the code. The full stack and its costs are broken down in the €3,250 stack post.

What would you do differently?

Validate before committing everything. I moved my life on three-year-old information and discovered the market had pivoted after I'd already burned the boats. That mistake is the reason validation comes before building in everything I teach — test the demand while it still costs you nothing.