Create Your First LLM App: Q&A with LangChain and Gemini

🚀 Your First Functional LangChain Project

Now that you understand LangChain's core components, it's time to build something real: a basic Q&A app.

This project will help you:

• Connect LangChain to a real LLM (Google Gemini)
• Pass user input into a prompt template
• Return structured answers to your frontend or CLI

🧠 What You'll Build

You'll create a workflow like this:

User input → Prompt template → Gemini response → Output answer

It's simple, powerful, and foundational for more complex chains later.

🛠️ Step-by-Step Setup

1. Install Required Packages

pip install langchain google-generativeai

Set your Gemini API key:

export GOOGLE_API_KEY="your-key"

2. Create the Chain

from langchain.llms import GooglePalm
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain

llm = GooglePalm(temperature=0.7)

prompt = PromptTemplate(
    input_variables=["question"],
    template="Answer the following question in 2–3 concise sentences:\n\n{question}"
)

chain = LLMChain(llm=llm, prompt=prompt)

response = chain.run("What is Retrieval-Augmented Generation?")
print(response)

✅ Why This Matters

This setup introduces key components:

• Prompt templating
• LLM response generation
• Clean chaining between input and output

It also gives you a starting point to integrate this into chatbots, APIs, or web frontends.

🔜 Next: Add Memory to Your App

In Chapter 3, we'll turn your stateless app into a conversation-ready chatbot using LangChain memory.

👉 Continue to Chapter 3: LangChain Memory & State — Building Context-Aware Conversations