Automation platforms are tools that connect individual applications and trigger sequences of steps without programming. For businesses they are an important complement to artificial intelligence, because the model on its own only generates an output – the saving only arises once the output automatically reaches where it belongs. A typical process works so that the trigger is a new message, form or document, the model processes the content, and the result is written into a system, sent by email or turned into a task. Deployment can be handled even by a non-technical person, but three rules apply to business use: the automation must have an owner, it must be documented, and it must have a defined response to what happens on error, otherwise the problem only surfaces after weeks of silent failure.
See also: AI use case, AI in Project Management, Shadow AI.
