Module 3 • Beginner

Types of AI Agents

Explore different AI agent architectures: Reactive, Deliberative, Hybrid, Single-Agent, and Multi-Agent systems with real-world examples.

📊 Module Progress

Complete objectives and quizzes to unlock your badge!
0/9 Tasks Completed

🎯 Learning Objectives

Differentiate between reactive, deliberative, and hybrid agents — Understand the core differences in architecture and use cases.
Understand single-agent vs multi-agent systems — Learn when to use one vs multiple specialized agents.
Identify which type of agent suits different tasks — Match agent architectures to real-world scenarios.
Recognize the JiXe ecosystem agent types — See how Hybrid and Multi-Agent systems work in practice.

📖 Theory

1. Reactive Agents

What: Respond to current input only, no memory of past interactions.
Example: Simple chatbot that answers FAQs.
Pros: Fast, simple, low resource usage.
Cons: Can't learn, can't handle complex tasks.

2. Deliberative Agents

What: Plan ahead, have memory, can reason.
Example: OpenClaw with memory enabled.
Pros: Handle complex tasks, learn from experience.
Cons: Slower, more resource-intensive.

3. Hybrid Agents

What: Combine reactive (fast responses) + deliberative (complex tasks).
Example: JiXe — quick chat + deep research when needed.
Pros: Best of both worlds.
Cons: More complex to build.

4. Single-Agent Systems

What: One agent handles everything.
Example: Basic OpenClaw setup (just JiXe).
Use case: Personal assistant for one user.

5. Multi-Agent Systems

What: Multiple specialized agents collaborate.
Example: Hermes (orchestrator) + JiXe + Kodi + Mark + Rei.
Use case: Complex projects requiring different expertise.
Benefit: Parallel processing, specialization.

Agent Types Comparison

Type Memory Planning Speed Best For
Reactive ❌ ❌ ⚡⚡⚡ Simple FAQs
Deliberative ✅ ✅ ⚡ Complex tasks
Hybrid ✅ ✅ ⚡⚡ All-around use
Multi-Agent ✅ ✅✅ ⚡⚡ Big projects

💡 Real-World Example: JiXe Ecosystem

🤖

Single-Agent Mode (Basic)

▼

User: "What's the weather in Ipoh?"
JiXe: Calls weather API → returns result
Simple, fast, one agent does everything

🏗️

Multi-Agent Mode (Advanced)

▼

User: "Build me a full Criticalyx AI course website."
Hermes (orchestrator):
- Spawns REI → Research competitor courses
- Spawns MARK → Write marketing copy
- Spawns KODI → Code the website
- Spawns DEX → Analyze user engagement data
Parallel processing, each agent specialized

🧪 Quick Check (5 Questions)

Q1: Which agent type has NO memory of past interactions?

A
Deliberative
B
Reactive
C
Hybrid
D
Multi-Agent

Q2: What is the main advantage of Multi-Agent systems?

A
They are simpler to build
B
They are faster for simple tasks
C
Parallel processing with specialization
D
They use less memory

Q3: Which type is OpenClaw with memory enabled?

A
Reactive
B
Deliberative
C
Hybrid
D
Single-Agent

Q4: What makes Hybrid agents special?

A
They have no memory
B
They only do simple tasks
C
They combine fast reactions with deep thinking
D
They use only one tool

Q5: In a Multi-Agent system, what is Hermes' role?

A
Code websites
B
Orchestrate and delegate tasks
C
Research information
D
Write marketing copy
← Previous
Module 2
Hub
🤖 AI Hub
🏗️
Badge Earned!
Agent Architect

You've completed all objectives and quizzes!