Lesson 1. What Is an AI Agent and How It Differs from a Chatbot#
Why This Matters#
The term "AI agent" sounds complex, but it's simply a program that understands a task, uses tools, and performs actions — like your assistant. Understanding the difference between a bot, an assistant, and an agent helps you choose the right solution and explain the value to clients.
Key Idea#
Agent = goal + context + tools + actions
A chatbot simply answers questions. An assistant helps and suggests. An agent actually does something: writes to a spreadsheet, sends notifications, analyzes data.
Understanding the Terms#
Chatbot:
-
answers questions based on a script or knowledge base
-
doesn't perform actions on its own
-
example: FAQ bot on a beauty salon website
AI Assistant:
-
helps with work: suggests, generates text, searches for information
-
general-purpose, not tailored to one task
-
example: ChatGPT, Claude
| ChatGPT | Claude |
|---|---|
AI Agent:
-
completes a specific task from start to finish
-
uses tools: spreadsheets, APIs, databases
-
can make decisions within given logic
-
example: lead qualification agent — asks questions, scores answers, writes results to CRM, and sends a notification to the manager
What an Agent Consists Of#
-
Goal — what needs to be done (e.g., qualify a lead)
-
Context — knowledge base, rules, constraints (e.g., pricing list and evaluation criteria)
-
Tools — what the agent can use (spreadsheets, forms, notifications)
-
Actions — what the agent does (asks questions, records answers, sends results)
Example: Agent vs Bot#
Task: handling a consultation request
Chatbot:
-
"Hello! Write your question, we'll reply within a day"
-
saves the message to a spreadsheet
-
manager figures it out themselves
AI Agent:
-
asks follow-up questions: "What's your goal?", "Do you have a budget?", "When do you need it?"
-
scores the lead by criteria (urgency, budget, readiness)
-
records everything in CRM with tags
-
sends the manager a notification: "Hot lead, budget ~$1,500, starting in a week"
Where Agents Are Used#
-
Customer support: handling common requests, preparing responses
-
Sales: lead qualification, preparing proposals
-
Marketing: review analysis, content selection
-
Operations: filling spreadsheets, data validation, reminders
-
HR: initial resume screening, candidate responses
Check Your Understanding#
-
How does an agent differ from a chatbot?
-
What are the components of an agent?
-
Give an example of a task that needs an agent, not just a bot
-
What tools can an agent use?