AI-first. Human-first.
These two ideas aren't in tension - they're the whole point. AI does the heavy lifting; humans decide, own, and stay in control. It's how we build One, and it's how we work every day, for ourselves and for our customers.
Our principles.
AI-first.
We reach for AI first, in the product and in our own work. Agents draft, summarize, research, and automate the heavy lifting so our people spend their time on judgment, taste, and the things only humans do well.
Human-first.
AI proposes; a human decides. Every consequential action, for us or for a customer, has a person in the loop who can see it, change it, and own it. We design for the human at the center, never the algorithm.
Owned, not rented.
The person owns their agent, their data, and the supercomputer it runs on. We build the same way internally: our tools answer to us, our data stays ours, and consent gates every exchange.
Consent is the default.
Nothing moves without permission. PCHP, consent receipts and a private transparency log, is how One works and how we expect to be treated. Restraint isn't a setting; it's the company.
How it shows up in practice.
We build with One.
We use our own agents to do our jobs, research, drafting, code, ops. If it isn't good enough for us, it isn't good enough for a customer.
Proof over promises.
We show the 60-second fetch, the working build, the real outcome, not adjectives. Claims wait for evidence, and for counsel where it matters.
Small teams, high ownership.
Founding-style ownership at every level. You own a number, a surface, or a customer, and the agents that help you scale it.
The human stays in control.
Every person owns their One Agent - it works with all their apps, APIs, and AI, seamlessly, with them in control and in ownership at all times. We hold ourselves to the same standard.
One Agent - it works with all their apps, APIs, and AI, seamlessly, with them in control and in ownership at all times. We hold ourselves to the same standard.