Robotic process automation (RPA): Robotic process automation (RPA) is software that mimics a person's clicks and keystrokes to run repetitive, rule-based tasks, such as copying invoice data from email into an accounting system. It follows a fixed script and makes no judgment calls.
Despite the name, there are no robots. An RPA "bot" is a script that opens applications, reads fields on screen, copies values, fills in forms, and clicks buttons in a fixed sequence, usually running on a virtual machine overnight or on a schedule. Platforms such as UiPath, Automation Anywhere, and Microsoft Power Automate let operations teams build these flows with a visual recorder rather than code.
RPA shines on high-volume, stable, boring work: moving data between two systems that do not talk to each other, reconciling reports, onboarding a new employee across five portals, processing standard claims. It is deterministic, which auditors like, and cheap per transaction. It is also brittle. Move a button, rename a field, or send it an invoice in an unexpected layout and the bot fails or, worse, keeps going with bad data.
That brittleness is where AI comes in. Vendors now bundle OCR, document understanding, and language models so bots can read messy inputs and handle exceptions, a combination often sold as intelligent automation. AI agents push further by pursuing a goal rather than a script, at the cost of predictability. The practical split: use RPA when the steps never change and correctness must be exact, use an agent when inputs vary and a human will check the result, and expect the two to keep blending inside the same platforms.
Example at work
An accounts payable team receives hundreds of supplier invoices a month as PDF attachments. An RPA bot pulls each attachment, hands it to a document-extraction model, enters the vendor, amount, and PO number into the ERP, and routes anything with a mismatched PO or a low confidence score to a clerk's exception queue. The clerks now touch only the exceptions.
Why it matters
RPA is the automation most mid-sized companies already own, and it is where AI is being bolted on first. Knowing what a bot can and cannot do helps you decide which processes to script, which to hand to an agent, and which still need a person.
Related terms
- Workflow automationWorkflow automation connects apps and steps so that a trigger, such as a new form response or an incoming email, automatically runs a sequence of actions, increasingly with an AI step that reads, decides, or writes along the way.
- Intelligent document processing (IDP)Intelligent document processing (IDP) uses OCR, machine learning, and language models to pull structured data out of documents such as invoices, contracts, and claims, validate it, and feed it into business systems, with humans reviewing the exceptions.
- Computer useComputer use is an AI capability in which a model operates a computer the way a person does: looking at the screen, moving the cursor, clicking, typing, and switching between apps and web pages to finish a task in software that has no direct integration.
- AI agentAn AI agent is a system that uses an AI model to pursue a goal on its own by planning steps, using tools such as web browsers, files, and software APIs, and adjusting based on results, rather than answering a single prompt.
- No-codeNo-code refers to tools that let people build apps, automations, and websites through visual interfaces and plain-language instructions instead of writing programming code.