AI Agent / Agentic AI
An AI agent is an LLM-driven system that, given a goal, plans a sequence of steps, calls tools or APIs, observes the results, and iterates until the goal is reached — going beyond single-turn chat to multi-step autonomous action.
Full definition
Modern agents combine an LLM as the planner/reasoner with tool use (function calling, code execution, web browsing, file operations), memory, and a control loop. Frameworks include OpenAI's Assistants and Responses APIs, Anthropic's Claude with tools, LangGraph, and Microsoft AutoGen. The 2024-2025 wave of "agentic AI" is moving from demos to production in coding (Cursor, Devin), customer support, research, and operations workflows.
Why it matters
Agentic AI is the next operating step beyond chat. Where a chatbot answers a question, an agent completes the task — drafting and sending the email, opening the ticket, running the analysis. Organizations deploying agents in 2025-2026 are reporting 30-70% cycle-time reductions on the workflows they target.
Example
An engineering team uses a coding agent to handle bug-fix tickets end-to-end: read the ticket, open the repo, run tests, reproduce the bug, write a fix, run tests again, open a pull request — with humans reviewing only the PR.
Related terms
- Large Language Model (LLM)A Large Language Model (LLM) is a deep neural network — almost always a transformer — trained on hundreds of billions to trillions of words to predict the next token, and to generate, summarize, translate, or reason over text.
- Tool Use (LLM)LLM tool use (also called function calling) is the ability of a language model to invoke external functions, APIs, databases, or systems based on the user's request — turning the LLM from a text generator into a controller that can act on the world.
- Chain-of-Thought (CoT)Chain-of-Thought (CoT) is a prompting and training technique in which a language model is encouraged to reason step-by-step before producing its final answer — improving accuracy on math, logic, and multi-hop reasoning by 10-40 percentage points on standard benchmarks.
- Context WindowA context window is the maximum amount of text — measured in tokens — a language model can read and reason over in one inference call, equivalent to the model's working memory for that turn.
- Retrieval-Augmented Generation (RAG)Retrieval-Augmented Generation (RAG) is a technique where a language model retrieves relevant documents from an external knowledge base before generating a response, reducing hallucination by 30-60% and enabling citation of sources.
Source & further reading
Primary source: Anthropic — "Building effective agents" (2024).
Citation policy: this entry is part of the AIDOLS AI Implementation Glossary and may be quoted for research, journalism, and education with attribution to aidolsgroup.com/it/glossary/ai-agent/.