Company
A small team that ships fast. AI-augmented engineering. 21-day prototyping cycles. Every idea gets built, tested, and either validated or killed. No permanent research projects.
Core Rule
Unless physically impossible due to the laws of physics, every hypothesis gets built and tested within 21 days. No exceptions. No "we need more research." No "let's think about it more." Build the thing, test the thing, learn from the thing.
The hypothesis was correct. The approach works. Move to the next stage — scale it, refine it, integrate it into the system.
The hypothesis was wrong. The approach doesn't work. Kill it fast. No sunk cost fallacy. No "just one more iteration." Move on to the next idea.
The results are inconclusive or suggest a modified approach. Refine the theory, formulate a new testable hypothesis, and start another 21-day cycle. Every cycle either converges on truth or eliminates a dead end.
AI-First
Every engineer has AI assistants for code, CAD, simulation, and analysis. Humans make decisions. AI handles the volume. This is how a small team outproduces companies 10x their size.
Custom simulation tools for testing robot behavior in virtual environments before committing to hardware. Thousands of scenarios tested per day. Fail in simulation, not in orbit.
AI-driven CAD tools for rapid mechanical design iteration. Generate, evaluate, and optimize designs faster than traditional workflows. From concept to printable design in hours, not weeks.
In-house 3D printing, CNC, electronics assembly. Design it today, hold it tomorrow. No external vendors. No 6-week lead times.
Culture
When in doubt, build. A bad prototype teaches more than a perfect plan. Meetings without outcomes don't happen here.
Stay small. Stay fast. Every person ships. No managers who don't build. AI amplifies each person's output by 10x.
Hardware has long feedback loops. The answer isn't to slow down — it's to shorten the loops. More prototypes, more tests, more data, faster.
Thermal vacuum chambers, vibration tables, radiation testing. Simulate space conditions on the ground. Validate before you fly.
Careers
Roboticists who've built real manipulators. AI engineers who've deployed autonomy systems. Aerospace engineers who've put hardware in orbit. Generalists who can do all three and don't need permission to move fast.
Founding team equity. San Francisco base. A problem hard enough to be interesting for the rest of your career. And robots that will actually fly — not sit in a lab.