What is an Intelligent automation?
Intelligent automation combines automation tools such as RPA and workflow engines with AI capabilities such as document understanding and language models, so processes that need reading or judgment can be automated too.
In more detail
The label describes a layered approach: rules and integrations for the predictable steps, AI for the unstructured ones, people for approvals and exceptions. It overlaps with hyperautomation, which stresses automating as many processes as possible across an organization, while intelligent automation names the technology mix. In practice the work is the same. Map the process, decide which steps need judgment, add an AI step only there and keep an exception route. A common failure is adding AI to a step that rules could handle, which costs more and is harder to test. Another is automating a process nobody has cleaned up. Providers of 'intelligent automation services' range from RPA integrators to AI-first teams, so ask for the architecture and the human checkpoints, since the label tells you little.
Example
A claims team keeps its RPA bot for copying data between two legacy screens, adds a language model to read adjuster notes and extract the cause of loss, and sends low-confidence extractions to a reviewer.