Follow the
hard questions.

For people who want to keep autonomous AI under control.

Conceptual porcelain arcs forming a collective sculpture with one cobalt segment

Who we want
to hear from.

  • Researchers in agent evaluation, AI security or human oversight.
  • Engineers who have built durable execution, permissions or observability.
  • People who have run AI agents in production and watched them fail.

No open roles listed right now. If your work fits, send it.

Introduce your work