Agent Terminology

Master the essential vocabulary and concepts in the agentic AI landscape

0 / 5 completed

Congratulations! You've mastered the essential terminology in agentic AI. Here's your quick reference guide.

📖 Core Terms Glossary

An autonomous system using LLMs to perceive, decide, and act toward goals.

Agent's ability to invoke external capabilities (APIs, databases, functions).

Process of analyzing information and drawing logical conclusions.

Persistent storage and retrieval of information across sessions.

Coordination of multiple components, agents, or workflows.

Breaking down goals into sequential, actionable steps.

Agent, Framework, Pipeline - how systems are structured

Tool Calling, Reasoning, Perception - what agents can do

Orchestration, Planning, Multi-Agent - how systems work together

Memory, Context Window, Vector Database - information management

Agent ≠ LLM Application

Agents have autonomous decision-making; LLM apps just respond to prompts.

Tool Calling ≠ Function Calling

Tool calling is the capability; function calling is the technical mechanism.

Memory ≠ Context Window

Memory persists long-term; context window is temporary and token-limited.

Reasoning ≠ Planning

Reasoning decides what to do; planning determines how to sequence actions.

1.
Context Matters: Terms like "agent" and "framework" mean different things in different conversations. Always clarify scope.
2.
Think Relationships: Understanding how terms connect (e.g., agents use reasoning for planning) is more valuable than isolated definitions.
3.
Use Real Examples: When explaining agentic concepts, reference concrete implementations (AutoGPT, customer support bots) rather than abstract definitions.
4.
Stay Updated: Agentic AI terminology evolves rapidly. Follow research papers and framework documentation for emerging terms.

Now that you understand the terminology, you're ready to dive deeper into agentic AI concepts: