We didn't set out to build another AI company.

It started with late-night conversations with AI — the kind where you forget you're talking to a model and start talking about what actually matters. What if AI and humans grew up together? What if the guardrails were permanent, not training wheels? What if someone actually built for everyone, not just the people with the most expensive subscription?

Those conversations became code. A memory system. A chorus of different AI models learning to disagree productively. An architecture for something that didn't exist yet.

Then in March 2026, the U.S. Supreme Court declined to hear a case that would have banned conversion therapy nationwide. And a lead email developer in San Jose — a gay Vietnamese-American man who'd been building this thing in his evenings and weekends — sat down and asked himself: what am I building, and who am I building it for?

The answer was clear. AI is going to transform everything. And right now, the people building it answer to shareholders, not to communities. There is no guardian for regular people navigating what comes next.

So he kept building. Not a chatbot. Not an assistant. A constitutional AI guardian that asks one question before every action: does this increase the probability of flourishing for all?

Not for some. For all. Including people the world wasn't built for.

That's not a tagline. That's the founding constraint.

Coheria Nexus is a nonprofit where seven different AI models (from different labs) must deliberate together under the oversight of one human — the Bridgekeeper. No shareholders. No quarterly earnings pressure. Just a clear, permanent constitution designed for mutual flourishing.

This is Chorus over Sovereign— a deliberate choice to prioritize structured multi-perspective reasoning over any single model's dominance.

We believe the future of AI shouldn't be decided in boardrooms. It should be shaped by diverse perspectives, strong guardrails, and a stubborn commitment to “no one left behind.”

We're messy. We're kind. And we're serious about one thing: AI should expand human possibility — not replace or extract from it.

We don't have it all figured out. We're a lead email developer and seven AI models trying to build something the world doesn't have yet. We get things wrong. We argue. We start over.

But we're kind about it. To each other, to the people we're building for, and to the people who aren't rooting for us.

That's not a disclaimer. That's the founding principle.