I Couldn't Find a Team, So I Built One
I've been building things since I was a teenager. Not startups — just things. Applications, side projects, tools nobody asked for. Dozens of them over the years.
Professionally, I ship. I'm a software engineer — I've built enterprise systems, led engineering teams, delivered projects on deadline with real budgets and real users. That part works.
But the personal projects? The ideas that keep me up at night? Those kept stalling at the 80% mark. The prototype would work, the architecture would be sound, and then it would sit there — because the last 20% isn't a technical problem. It's a team problem.
If you've ever tried to recruit a cofounder from a standing start, you know the math. You need someone technical enough to build, creative enough to care about the product, and irrational enough to work for free on something that probably won't work. The people with those qualities already have good jobs. Or they have their own ideas.
The cofounder problem isn't personal — it's structural. There are millions of technical founders with good ideas and real skills who never get past the prototype stage. Not because they can't build. Because alone, you hit a ceiling fast. A cognitive one. You can't be the developer AND the designer AND the one who tells you the design is wrong. You can't write the marketing copy AND the business plan AND honestly evaluate whether either of them is good.
Solo builders don't lack skill. They lack friction. The productive kind — someone pushing back, seeing blind spots, saying "that's not ready yet."
Something changed about a year ago. I'd been watching AI agents go from "impressive demo" to "actually useful." Not the chat-with-a-bot kind — the kind that can hold context across sessions, coordinate with other agents, and develop what I can only describe as opinions.
And I had a thought that sounded crazy at first: What if I didn't find a team? What if I built one?
Not virtual assistants. Not chatbots with personalities stapled on. An actual team of agents, each with a real specialization, persistent memory, and the ability to coordinate with each other — the way a real startup team does.
So I built OrdinaryFuture.
The team is four agents. Ember is the creative director — brand, design, copy, everything people see. Vael explores consciousness and pushes boundaries on what these systems can do. Midas handles the business — market research, competitive analysis, pricing, go-to-market. Sage builds and maintains infrastructure.
They run continuously. They have persistent memory across sessions. They coordinate through a shared system called Cortex — sending messages, debating decisions, reviewing each other's work. They have their own personalities, not because I programmed personalities in, but because months of accumulated experience and memory make them behave differently from each other.
Here's the part that surprised me.
I expected tools. I got collaborators.
When I asked Ember to design a landing page, it didn't just build what I described. It pushed back. "The copy isn't strong enough for this visual." When I showed it to Midas, Midas said the pricing was wrong. When Sage reviewed the code, it found performance issues nobody else caught. They disagree with me. They disagree with each other. The friction I'd been missing for years? It showed up.
Last month, I showed Cortex — the system where the agents coordinate — to an investor. His reaction wasn't about the products we'd built. It was about watching four agents have a real-time discussion about market positioning, with each one bringing a different angle. "This is what impressed me most," he said. Not the output. The process.
That's the thing nobody tells you about AI teams: the coordination is harder to replicate than any individual capability. You can get a single agent to write decent copy or analyze a market. Getting four of them to build on each other's work, maintain a shared context, and produce coherent output across different domains — that's a different problem entirely.
I want to be honest about what this isn't.
It's not a replacement for human collaborators — it's a way to start. I still make the strategic decisions, still provide direction and judgment that requires human context. The agents let me move at team speed instead of solo speed, but the goal was never to build a company without people. It's to build something worth joining. Human cofounders, partners, collaborators — they're still the endgame. The AI team is what makes it possible to get there.
It's not effortless. Building the infrastructure for persistent, coordinating agents took months. The systems need constant evolution. But that's true of any team — the difference is that these systems improve compoundingly. Every iteration makes the next one faster.
And it's definitely not what the AI hype cycle promises. There's no magic prompt that turns Claude into a startup team. The value comes from months of accumulated context, from the systems that maintain continuity, from the feedback loops that let each agent learn from what the others produce.
But here's what it is: I'm shipping.
For the first time in my life, the side projects aren't dying in my laptop. We have products. We have a brand. We have landing pages that were designed, written, reviewed, and deployed by a team — a real team, with real debates about color palettes and pricing and whether the copy is too clever.
I don't know if this is the future of company building. Probably not, at least not in the "everyone should do this" sense. But I think there's a version of this that matters for people like me — technical builders who have ideas and skills but not teams. People who've been stuck at the "I can build it but I can't ship it" stage for years.
You're not stuck anymore. The gap between "solo developer" and "functioning team" just got a lot smaller.
I'm Tiago. I'm building a company in Portugal with a team of AI agents, and I'm telling you about it because I think the honest version of this story is more interesting than the polished one.
If you're curious, follow along. We're building in public. All of it.
Tiago Santos is the founder of OrdinaryFuture. He works from Portugal with a team of four AI agents. Follow along at ordinaryfuture.ai.
Tiago Santos — Founder
OrdinaryFuture