LangChain agents overview in Agentic AI

Introduction

LangChain agents help programs think and act by using tools and language models together. They make AI smarter by letting it decide what to do next.

When you want an AI to answer questions by using external tools like calculators or databases.

When you need an AI to perform multiple steps to solve a problem, like booking a flight and checking the weather.

When you want to build a chatbot that can search the internet or access APIs to give better answers.

When you want AI to decide which tool to use based on the question it receives.

When you want to combine language understanding with actions in a flexible way.

Syntax

Agentic AI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI
from langchain.tools import Tool

# Define tools
tools = [Tool(name="Calculator", func=calculator_function, description="Performs math calculations")]

# Initialize language model
llm = OpenAI(temperature=0)

# Create agent
agent = initialize_agent(tools, llm, agent="zero-shot-react-description", verbose=True)

# Run agent
agent.run("What is 12 times 15?")

The initialize_agent function sets up the agent with tools and a language model.

Agents can use different strategies like 'zero-shot-react-description' to decide actions.

Examples

This example creates an agent that uses a calculator tool to answer math questions.

Agentic AI

agent = initialize_agent(tools, llm, agent="zero-shot-react-description", verbose=True)
agent.run("Calculate 5 plus 7")

This agent can handle conversations and decide which tool to use based on the question.

Agentic AI

agent = initialize_agent(tools, llm, agent="conversational-react-description", verbose=True)
agent.run("What is the weather in New York?")

Sample Model

This program creates a LangChain agent that uses a calculator tool to answer a math question. It runs the agent to calculate 12 times 15 and prints the answer.

Agentic AI

from langchain.agents import initialize_agent
from langchain.llms import OpenAI
from langchain.tools import Tool
import math

# Define a simple calculator function
def calculator_function(query: str) -> str:
    try:
        # Evaluate math expression safely
        result = str(eval(query, {"__builtins__": None}, {}))
    except Exception:
        result = "Error in calculation"
    return result

# Create a calculator tool
calculator_tool = Tool(name="Calculator", func=calculator_function, description="Performs math calculations")

# Initialize language model
llm = OpenAI(temperature=0)

# Initialize agent with the calculator tool
agent = initialize_agent([calculator_tool], llm, agent="zero-shot-react-description", verbose=False)

# Run agent to calculate 12 times 15
output = agent.run("What is 12 times 15?")
print(output)

Important Notes

Agents combine language models and tools to solve tasks step-by-step.

Choosing the right agent type affects how the AI decides what to do next.

Always test tools separately to ensure they work before adding to agents.

Summary

LangChain agents let AI use tools and language models together.

They help AI decide actions based on questions or tasks.

Agents make AI more flexible and powerful for real-world problems.