Structured chat agent in LangChain - Mini Project: Build & Apply

Build a Structured Chat Agent with LangChain

📖 Scenario: You want to create a simple chat agent that can answer questions about a specific topic using LangChain. This agent will have a clear structure to handle user input and provide relevant answers.

🎯 Goal: Build a structured chat agent using LangChain that can receive a question and return a relevant answer based on a predefined knowledge base.

📋 What You'll Learn

Create a knowledge base as a dictionary with exact entries

Set up a LangChain prompt template for the chat agent

Implement the core logic to process user input and generate answers

Complete the agent setup to run the chat interaction

💡 Why This Matters

🌍 Real World

Structured chat agents help businesses provide quick, accurate answers to customer questions using predefined knowledge.

💼 Career

Understanding how to build chat agents is useful for roles in AI development, customer support automation, and software engineering.

1

Create the knowledge base dictionary

Create a dictionary called knowledge_base with these exact entries: 'Python': 'A popular programming language.', 'LangChain': 'A framework for building language model applications.', 'Agent': 'An entity that can perform tasks based on input.'

LangChain

# Your code here

Need a hint?

Use a Python dictionary with the exact keys and values given.

2

Set up the prompt template

Create a variable called prompt_template and assign it the string: 'Answer the question based on the knowledge base: {question}'

LangChain

knowledge_base = {
    'Python': 'A popular programming language.',
    'LangChain': 'A framework for building language model applications.',
    'Agent': 'An entity that can perform tasks based on input.'
}

# Your code here

Need a hint?

Assign the exact string to prompt_template.

3

Implement the core logic to answer questions

Define a function called answer_question that takes a parameter question. Inside, use a for loop with variables key and value to iterate over knowledge_base.items(). If key is found in question, return value. If no key matches, return 'Sorry, I do not know the answer.'

LangChain

knowledge_base = {
    'Python': 'A popular programming language.',
    'LangChain': 'A framework for building language model applications.',
    'Agent': 'An entity that can perform tasks based on input.'
}

prompt_template = 'Answer the question based on the knowledge base: {question}'

# Your code here

Need a hint?

Use a function with a for loop to check if any key is in the question string.

4

Complete the chat agent setup

Create a variable called chat_agent and assign it a dictionary with two keys: 'prompt' set to prompt_template and 'answer_function' set to answer_question

LangChain

knowledge_base = {
    'Python': 'A popular programming language.',
    'LangChain': 'A framework for building language model applications.',
    'Agent': 'An entity that can perform tasks based on input.'
}

prompt_template = 'Answer the question based on the knowledge base: {question}'

def answer_question(question):
    for key, value in knowledge_base.items():
        if key in question:
            return value
    return 'Sorry, I do not know the answer.'

# Your code here

Need a hint?

Create a dictionary named chat_agent with the specified keys and values.