Skip to content
Use cases
Autopilotplanned
Autopilotplanned

Hold spend under a cap without watching it

A policy that acts on its own, with every action written into the same activity log as everything else, and approval required before it applies.

Everything below describes what this is being built to do. It needs Autopilot, which you can look around today but cannot yet use for the whole of this job. The problems are real now; the answers are the part still being built.

A budget that needs somebody watching it is not a budget.

Why it is hard

  • You find out at the invoice

    Spend is visible monthly and decided by the second, which is the wrong way round for anybody who has to answer for it.

  • A hard stop destroys work

    The blunt fix is killing everything at a threshold, which turns a budget overrun into lost work and usually costs more than it saved.

  • Automatic changes are a trust problem

    A tool that alters your infrastructure while you are asleep is one incident away from being switched off entirely.

What answers it

  • A limit that refuses rather than destroys

    Reaching it refuses new rentals while running ones bill to completion. That distinction already exists in the product today and is deliberate.

  • Approval is a step, not a setting

    Automatic application is opt-in per policy and off by default. Nothing moves money or hardware without somebody saying so.

  • Every action is written down

    Whatever a policy does lands in the same activity log as everything else, in euro, rather than in a private audit nobody can read.

Spend limits work today. Policies acting on their own are the part still being built.