The words behind an AI employee.
Short definitions for the way we talk about the work, the loop, and the controls that keep it safe.
- AI employee
- A model plus the operating system it needs to hold a job, run by a human operator to an outcome. Not a chatbot and not a tool you babysit, a role that delivers work.
- Blueprint
- The homepage tool that researches your company and returns up to three grounded AI-employee proposals, each one only shown if it clears a strict review.
- Guardrails
- The rulebook, sign-off gate, kill switch, spend cap, sandbox, and traces together, the safety system that lets an operator run on real work.
- Horizon move
- A next-stage capability most companies plan a year or two out, stood up now because the signals say it is time.
- Kill switch
- The control that stops an operator instantly. Operators fail closed, never open.
- Operator
- The forward-deployed human who runs the loop and owns the result. Every AI employee has one on top.
- Rulebook
- The lines an operator never crosses, set before it runs.
- Runbook
- The written steps for the work an operator does over and over, so the same job is done the same way every time.
- Sense, act, check, learn
- The four steps of the loop. Sense what is happening, act on it, check with a person where judgment is needed, and learn so the next run is sharper.
- Sign-off gate
- The point in the loop where a human approves anything that carries risk before it goes out.
- Spend cap
- A hard limit on what an operator can spend, so it can never run up a bill.
- The loop
- The single signature device behind every operator: sense the trigger, act on the work, check with a human on anything risky, and learn from what worked.