Build at the speed of your judgment

You own a mission, not a list of tasks

You will not wait for a founder to keep your backlog full. We will give you direct access to the context the company runs on: what we believe, what we are building, what matters now, what customers need, and what good looks like.

From there, you own forward motion.

That might mean breaking a product goal into parallel lines of research, design, implementation, and testing. It might mean discovering that the original plan was wrong and bringing back a better one. It might mean slowing down for one consequential decision before accelerating everything downstream of it.

Autonomy here does not mean working alone or inventing your own direction. It means having enough context to act without waiting for someone to translate the mission into tasks.

You will move unusually fast

An idea can become a working interface while it is still fresh enough to argue about. A reference that would normally trigger weeks of estimation can be taken apart, explored, implemented, and tested in days. Product, design, engineering, and quality do not have to wait in a line when one person can keep several useful loops moving.

Speed only matters when the result holds up. You are responsible for the checks, feedback, and review that make your output trustworthy. We want people who can move quickly because they built a better loop, not because they skipped the work required to know whether something is good.

Every project should leave two things behind: a product improvement and a better system for producing the next one.

Your range will expand

Agents make technical boundaries cheaper to cross. Your title will not define the edge of what you can own.

An engineer can explore interaction design, run a visual teardown, build several variants, automate browser testing, and improve the backend contract underneath the winning version. A designer can make feedback more systematic, test ideas in the real product, and turn repeated judgment into a reusable standard.

Strong foundations still matter. Taste matters more. So do curiosity, initiative, and the ability to tell when generated work is wrong. We care less about whether you have spent five years in the exact layer a problem touches and more about whether you can learn what the mission requires and deliver the whole result.

You will shape how this kind of company works

We are not only using this operating model. We are building it into the way we work: a system where shared context and goals become agent work, and every piece of that work is checked against what the company believes, in writing. We run on that system every day, and it is becoming part of what we ship.

Joining now means helping define both the products and the way an AI-native company operates. The workflows you improve for yourself can become capabilities other companies use to move faster. Few roles let you work on the product, the company building it, and the operating model behind both at the same time.

We work together in San Francisco

We strongly prefer to build this team in San Francisco. In-person work gives a small group more shared context, faster feedback, and more room to push an idea around until it becomes something better.

We will consider an exceptional remote candidate who has already shown that they can create direction from context, communicate clearly in writing, and keep important work moving without synchronous coordination.

You will probably thrive here if

• you have taken a rough goal to a shipped result without waiting for a complete task list
• you see an ambitious reference and immediately want to understand how to make it real
• you already use agents to expand what you can own, and you keep improving that system
• you care about the experience and the implementation, not only whether the feature technically works
• you like being close to the mission, the customer, and the consequences of your decisions
• you want your rate of learning and responsibility to grow much faster than your title normally would

This will probably be frustrating if you want a stable backlog, a narrow layer to own, or a manager who decides what you should do next. Those are reasonable ways to work. They are not how this company will win.