MalleableMalleable
← All features
Your team changes it

Self-improving operations

Malleable suggests improvements from real runs. Your team approves each change.

A review notices a pattern across real runs and proposes a change. It goes in only once a person approves it.

How it works

  • Regular reviewsMalleable reviews each operation on a schedule, and soon after a run fails.
  • It learns from real runsIt looks for steps that break, work that's wasted, and problems the AI had to work around.
  • Your team approves each changeSuggestions come in plain language. Approve, dismiss or snooze each one. Nothing changes until you approve.
  • Changes are tested firstAn approved change goes into a draft you can test before it goes live.
Three suggestions in plain language, each with a checkbox, and buttons to start working on them or wait
Suggestions in the editor, in plain language. Pick the ones to make.

For example

After a few weeks of runs, the review notices that new customers are often asked for the same tax form twice. It suggests asking once, at the start. The owner approves the change, and later runs ask only once.

More detail in the docs: Suggestions →

Book a demo