What is intelligent automation?
Intelligent automation combines rule-based automation, such as workflow software and robotic process automation (RPA) bots, with AI models that handle the steps rules cannot. Common AI components include intelligent document processing to read forms and letters, text classification to sort incoming requests, and predictive models to score risk or priority. The workflow still defines the path. AI turns messy inputs into structured data, bots and integrations act on that data, and people review whatever the models are unsure about.
The term overlaps with hyperautomation, which describes an organization-wide program of automating as much work as possible with a mix of these tools. Agentic process automation goes a step further by letting agents choose the next step as well.
Intelligent automation in practice
A specialty clinic group receives hundreds of faxed referrals a day. A document model reads each fax and extracts the patient, the referring provider, and the diagnosis, with a confidence score for each field. A bot creates the referral in the practice management system and checks insurance eligibility. Referrals with a low-confidence field or a missing diagnosis go to an intake coordinator, who sees the fax beside the extracted data and corrects it. Those corrections become training data for the next version of the document model. Coordinators spend their time on the problem referrals, and complete ones reach scheduling the same day.