There's a specific kind of startup that announces itself with a splash page, a waitlist, and six months of silence. Then it either ships something mid, or quietly disappears.
We decided early that Orochi wouldn't work that way.
The Default is Opacity
Most companies default to secrecy. The reasoning sounds rational: competitive advantage, premature expectations, the fear of showing unfinished work.
But secrecy has costs that rarely make the spreadsheet:
- Internal decisions go unchallenged because there's no external pressure to justify them
- Product-market fit assumptions calcify without real feedback loops
- The team optimizes for internal consensus instead of external value
Transparency inverts these dynamics.
What Building in Public Actually Means
It doesn't mean livestreaming your standups or tweeting every commit. That's performance, not transparency.
For us, it means:
Sharing the reasoning, not just the results. When we make an architectural decision, we explain why. When we change direction, we explain what we learned. The thinking is more valuable than the doing.
Admitting what we don't know. Every product has assumptions baked in. We'd rather name ours explicitly than pretend they don't exist. "We think this will work because X, and we'll know we're wrong if Y" is more honest than "Our revolutionary AI will transform..."
Making our values falsifiable. Anyone can claim to care about user privacy. Building in public means people can check. If our code doesn't match our claims, someone will notice — and that's the point.
The Uncomfortable Part
Building in public is uncomfortable because it removes the buffer between intention and execution. You can't say "we prioritize quality" while shipping broken features. You can't say "we value user feedback" while ignoring bug reports.
The discomfort is a feature. It's the mechanism that keeps the gap between what you say and what you do from growing too wide.
Why This Blog Exists
This blog is part of that commitment. Not every post will be profound. Some will be technical deep-dives. Some will be half-formed thoughts about where AI is headed. Some will age badly.
That's fine. The point isn't to be right about everything — it's to show the work, including the parts that are messy and uncertain.
If you're interested in following along, you're in the right place. If you think we're wrong about something, even better — that's how the work gets better.