# Agentic AI

> Agentic AI describes AI systems that pursue a goal across multiple steps instead of answering a single prompt. An agentic system plans, calls tools such as APIs or databases, checks the results and decides what to do next, usually within limits set by its builders. The term covers a spectrum, from a workflow with one AI decision point to a fully autonomous agent.

## Why it matters for your business

Most vendor pitches now use the word agentic, and it can mean anything from a chatbot with one API call to software that acts on your systems unsupervised. For a buyer, the useful questions are concrete: which actions can the system take, which require approval, what happens when it is unsure, and how is each step logged. The more autonomy a system has, the more it needs permissions, testing and monitoring around it. Many business processes get most of the value from a mostly fixed workflow with a few AI decisions inside it.

## Example

An accounts payable team receives invoices by email. An agentic workflow reads each invoice, looks up the vendor and purchase order in the accounting system, flags mismatches and drafts the bill for approval. It decides on its own whether to ask the vendor for a missing PO number, but a person approves every payment before it is released.

Related terms: https://www.lumeor.studio/glossary/ai-agent/, https://www.lumeor.studio/glossary/custom-ai-agents/, https://www.lumeor.studio/glossary/tool-calling/, https://www.lumeor.studio/glossary/human-in-the-loop/, https://www.lumeor.studio/glossary/multi-agent-system/

---

Source: https://www.lumeor.studio/glossary/agentic-ai/