tyeetaleTwo years of solo building. 100 daily active users. A product people genuinely loved. And I shut it down.
The business model wasn’t working. Affiliate revenue couldn’t sustain growth, direct charging was a hard sell, and I was one person doing everything: engineering, design, user research, marketing, support. The compounding wasn’t on the revenue side. It was on the learning side.
I’d learned how to build from zero. What I hadn’t learned was how to build at scale. Real scale. Not “100 users” scale. Billion-dollar-infrastructure scale.
Blue Origin was the answer to a specific question I had: what does data infrastructure look like when the stakes are a $5.4 billion rocket program? When the data has to be right because actual hardware decisions depend on it? When 320 budget owners need to make financial decisions from your pipelines?
The difference between startup data and enterprise data isn’t complexity. It’s consequence. In a startup, if your data is wrong, you make a bad product decision. At Blue Origin, if your data is wrong, someone might make a bad call on a rocket subsystem.
That stakes calibration changed how I think about systems. Every pipeline I build now, I ask: what’s the cost of this being wrong? Who’s making decisions from this? How do I make it trustworthy at the moment it matters?
I don’t regret the two years of solo building. They taught me to move fast and make decisions alone. But there are things you can only learn inside a system that’s bigger than you. The combination of both is what makes the next thing possible.