Case Studies in AI Agents

Duration: 5 min

This module delves into various case studies of AI agents, exploring their functionalities, applications, and the underlying technologies that enable them. Understanding these case studies is crucial for grasping the potential and limitations of AI agents in real-world scenarios.

ReAct Framework

The ReAct (Reason and Act) framework allows AI agents to reason about a task and then act upon it. This involves generating a plan, executing actions, and observing the outcomes to refine future actions. It is particularly useful in dynamic environments where conditions change frequently.

import random

def react_agent():
    # Define a simple task: find a random number between 1 and 10
    target = random.randint(1, 10)
    current = 0
    attempts = 0
    
    while current!= target:
        # Reason: Choose a new number
        current = random.randint(1, 10)
        # Act: Increment attempts
        attempts += 1
        print(f'Attempt {attempts}: Current number is {current}')
    
    print(f'Target {target} found in {attempts} attempts!')

# Run the ReAct agent
react_agent()

Try it in Google Colab:

Attempt 1: Current number is 3
Attempt 2: Current number is 7
Attempt 3: Current number is 5
Attempt 4: Current number is 9
Target 9 found in 4 attempts!

LangGraph for Multi-Agent Systems

LangGraph is a framework for creating multi-agent systems where multiple AI agents collaborate to solve complex problems. Each agent can have its own goals and actions, and they communicate to achieve a common objective. This is particularly useful in scenarios requiring coordinated efforts among different entities.

from langgraph import LangGraph

# Define two simple agents
def agent1():
    return 'Agent 1: Hello!'

def agent2():
    return 'Agent 2: Hi there!'

# Create a LangGraph instance
graph = LangGraph()

# Add agents to the graph
graph.add_agent('agent1', agent1)
graph.add_agent('agent2', agent2)

# Define interactions
graph.add_interaction('agent1', 'agent2')

# Run the multi-agent system
output = graph.run()
print(output)

💡 Tip: Ensure that agents in a multi-agent system have clear, non-overlapping goals to avoid conflicts and enhance collaboration efficiency.

Key Concepts

Check Your Understanding