Hey there! As I venture into building agentic MEAN apps with LangChain.js, I wanted to take a step back and revisit the core concepts of the MEAN stack. LangChain.js brings AI-powered automation and reasoning capabilities, enabling the development of agentic AI applications such as intelligent chatbots, automated customer support systems, AI-driven recommendation engines, and data analysis pipelines. Understanding how it integrates into the MEAN stack is essential for leveraging its full potential in creating these advanced applications. So, I put together this quick learning blog to share what I’ve revisited.
The MEAN stack is a popular full-stack JavaScript framework that consists of MongoDB, Express.js, Angular, and Node.js. Each component plays a crucial role in developing a seamless web application. Additionally, LangChain.js can be integrated within the MEAN stack to introduce intelligent workflows, AI-driven automation, and enhanced data processing capabilities. In this blog, we’ll break down the MEAN stack architecture and explain how these technologies work together while identifying where LangChain.js fits in.
Breaking Down the MEAN Stack
1. Angular – The Frontend
Angular is a powerful front-end framework used to create dynamic and responsive user interfaces. It runs in the browser and is responsible for rendering UI components and managing user interactions.
Key Features:
- Handles routing and navigation within the application.
- Makes HTTP requests to the backend via APIs.
- Provides a component-based architecture for scalability.
- Uses services to fetch and manipulate data.
Setting up Angular using Angular-CLI
To create an Angular project using WebStorm IDE, follow these steps:
- Open WebStorm and click on Create New Project.
- Select Angular CLI from the list of available project types.
- Enter a project name and choose a location to save the project.
- Click Create and wait for WebStorm to initialize the Angular project.
- Open the Terminal in WebStorm and run the following command to serve the application:
ng serve --open
Modify app.component.ts to Test with Express Server
Create or update the app.component.ts file with the following code:
import { Component, OnInit } from '@angular/core';
import { ApiService } from './services/api.service';
import {FormsModule} from '@angular/forms';
import {CommonModule} from '@angular/common';
@Component({
selector: 'app-root',
imports: [FormsModule, CommonModule],
templateUrl: './app.component.html',
styleUrls: ['./app.component.css']
})
export class AppComponent implements OnInit {
userInput: string = ''; // Holds user input
responseMessage: string = ''; // Holds AI response
isLoading: boolean = false; // Show loading state
constructor(private apiService: ApiService) { }
ngOnInit(): void {
}
// Function to send user input to the backend
askQuestion() {
if (!this.userInput.trim()) {
this.responseMessage = 'Please enter a question.';
return;
}
this.isLoading = true; // Show loading indicator
this.responseMessage = '';
this.apiService.askQuery(this.userInput).subscribe(
response => {
this.responseMessage = response.answer; // Extract 'answer' field from JSON
this.isLoading = false;
},
error => {
console.error('Error fetching response:', error);
this.responseMessage = 'Error fetching data. Please try again.';
this.isLoading = false;
}
);
}
}
Modify app.component.html file
Add the following code in app.component.html file
<div style="text-align: center; margin-top: 50px;">
<h1>Ask AI a Question</h1>
<input type="text" [(ngModel)]="userInput" placeholder="Enter your question..."
style="width: 60%; padding: 10px; margin: 10px; border-radius: 5px; border: 1px solid #ccc;">
<button (click)="askQuestion()"
style="padding: 10px 20px; background-color: #007BFF; color: white; border: none; border-radius: 5px; cursor: pointer;">
Ask
</button>
<p *ngIf="isLoading" style="color: gray; font-style: italic;">Loading...</p>
<h2 *ngIf="responseMessage">{{ responseMessage }}</h2>
</div>
Create a Service File (app.service.ts)
To enable communication between the Angular frontend and the Express backend, we need to create an api.service.ts file within the src/app/services folder. This service handles HTTP requests to the backend API endpoints.
The apiUrl in the api.service.ts file is set to http://localhost:5000/users. This URL represents the endpoint exposed on the Express server. When the Angular frontend sends a request to this URL, it is received by the users.js file in the Express backend. The server then processes the request and returns a JSON response, which Angular consumes and displays in the UI.
import { Injectable } from '@angular/core';
import { HttpClient } from '@angular/common/http';
import { Observable } from 'rxjs';
@Injectable({
providedIn: 'root'
})
export class ApiService {
private apiUrl = 'http://localhost:5000/users';
private queryllmUrl = 'http://localhost:5000/queryllm';
constructor(private http: HttpClient) { }
getMessage(): Observable<any> { // Expect JSON, not text
return this.http.get<any>(this.apiUrl);
}
// Function to send user input to the backend
askQuery(question: string): Observable<any> {
return this.http.post<any>(this.queryllmUrl, { question });
}
}
Configure app.config.ts
Here is the code for app.config.ts:
import { ApplicationConfig, provideZoneChangeDetection } from '@angular/core';
import { provideRouter } from '@angular/router';
import { routes } from './app.routes';
import { provideHttpClient } from '@angular/common/http';
export const appConfig: ApplicationConfig = {
providers: [
provideZoneChangeDetection({ eventCoalescing: true }),
provideRouter(routes),
provideHttpClient()
]
};
Where LangChain.js fits: If you want to incorporate AI-driven suggestions or chatbot interactions in your Angular app, you can make API calls to a LangChain-powered backend (express server) that processes natural language inputs and returns meaningful responses.
2. Express – The Backend Framework
Express.js is a minimal and flexible Node.js web framework that is used to build APIs and manage backend logic. It runs on the server and serves as the communication layer between Angular and MongoDB.
When creating a new Express project using WebStorm IDE, the default setup includes two key files: index.js and users.js. The index.js file serves as the main entry point, while users.js handles user-related API requests. For working with LLM, I added third file under routes folder, namely, queryllm.js. Below is an example of how queryllm.js can be structured to generate text response using OpenAI LLM and send back the response to the Angular frontend. Additionally, ensure that the following code exists in app.js on the Express server to properly route requests to queryllm.js:
var queryllmRouter = require("./routes/queryllm");
app.use("/queryllm", queryllmRouter);
This configuration allows Express to correctly handle requests sent to http://localhost:5000/queryllm by forwarding them to users.js. The response from queryllm.js is then sent back to the Angular frontend. The queryllm.js file is found within routes folder.
var express = require("express");
var router = express.Router();
require("dotenv").config();
const { ChatOpenAI } = require("@langchain/openai"); // ✅ Use require instead of import
const model = new ChatOpenAI({
model: "gpt-4o",
temperature: 0, // Adjust as needed
openAIApiKey: process.env.OPENAI_API_KEY,
});
/* POST request to get AI response */
router.post("/", async function (req, res) {
try {
const { question } = req.body;
if (!question) {
return res.status(400).json({ error: "Question is required" });
}
const response = await model.invoke(question);
res.json({ answer: response.text });
} catch (error) {
console.error("Error calling OpenAI:", error);
res.status(500).json({ error: "Failed to process request" });
}
});
module.exports = router;
In order for above code to work, make sure to add OPENAI_API_KEY parameter value in “.env” file.
Key Features:
- Handles API routing and middleware functions.
- Processes HTTP requests (GET, POST, PUT, DELETE).
- Manages authentication and session handling.
- Works seamlessly with Node.js to serve dynamic content.
Where LangChain.js fits: LangChain.js can be integrated into Express app thereby creating LLM / Retrieval Augmented Generation (RAG) and agentic AI apps. For example, you can have an Express route that leverages LangChain.js to process user queries and generate responses dynamically.
3. MongoDB – The Database
Once the Express app is created, the following package needs to be installed to interact with MongoDB:
npm install mongoose cors dotenv
Next, create a database in MongoDB. This can be done using a MongoDB client such as MongoDB Compass.
After setting up the database, add the following environment variables in a .env file in the root folder of the Express app:
MONGO_URI=mongodb://localhost:27017/hellolc
PORT=5000
- hellolc is the database name in the MongoDB instance running on port 27017.
- 5000 is the port on which the Express server is running.
MongoDB is a NoSQL database that stores data in a document-based format (JSON-like structure). It is highly flexible, scalable, and integrates well with JavaScript applications.
Ensure that the following code exists in app.js to establish a connection between the Express app and MongoDB:
const mongoose = require('mongoose');
require('dotenv').config();
mongoose.connect(process.env.MONGO_URI, {
useNewUrlParser: true,
useUnifiedTopology: true
}).then(() => console.log('MongoDB Connected'))
.catch(err => console.log(err));
This code reads the MongoDB connection string from the .env file and establishes a connection using Mongoose. If the connection is successful, a message MongoDDB Connected will be logged to the console; otherwise, the error will be logged.
MongoDB is a NoSQL database that stores data in a document-based format (JSON-like structure). It is highly flexible, scalable, and integrates well with JavaScript applications.
Key Features:
- Schema-less database structure for flexible data storage.
- Stores and retrieves JSON documents efficiently.
- Provides high scalability and real-time performance.
- Supports indexing and advanced query mechanisms.
Where LangChain.js fits: If you’re working with unstructured or AI-generated data, LangChain.js can store conversation history, embeddings, and metadata in MongoDB for efficient retrieval and further processing.
4. Node.js – The Backbone of the MEAN Stack
Node.js is a JavaScript runtime environment that allows developers to run JavaScript code outside the browser. It serves as the backbone of the MEAN stack by running the Express server and connecting the frontend (Angular) with the backend (MongoDB).
Key Features:
- Uses an event-driven, non-blocking I/O model for high efficiency.
- Runs the Express server to handle API requests.
- Connects to MongoDB and retrieves data asynchronously.
- Manages dependencies using npm (Node Package Manager).
Where LangChain.js fits: Node.js can run LangChain.js workflows to integrate AI models, execute reasoning chains, and process user inputs. By setting up LangChain.js in a Node.js backend, your application can include smart decision-making and conversational AI capabilities.
Check out my latest book titled as First Principles Thinking: Building winning products using first principles thinking.
- How to Setup MEAN App with LangChain.js - February 9, 2025
- Build AI Chatbots for SAAS Using LLMs, RAG, Multi-Agent Frameworks - February 8, 2025
- Creating a RAG Application Using LangGraph: Example Code - February 1, 2025
I found it very helpful. However the differences are not too understandable for me