Tech Stack Impact Explorer

Tech Stack Basics

  • Three layers make almost every website and app work: the front-end, the service layer, and the database. Learn what they are, why you need all three, how each one is programmed, where each one sits — and how a click travels through them.

Three layers, not one screen

What you see in the browser looks like one thing. It is not. Behind every profile, cart, or filter are three jobs that have to live in three different places: draw the screen, decide what is allowed, and remember the data.

Those jobs are the front-end, the service layer (back-end), and the database (also back-end). Skip one, and the product either cannot show anything, cannot save anything, or cannot stay safe.

The three layers

Same three names you will see on Popular tech stack options and in the explorer. Here is what each one is for.

01 · What you see

Front-end

The front-end defines the look and layout of pages, menus, submenus, buttons, and dropdowns, plus how they respond to clicks. On DuckBuddy that is Anna’s profile, the live chat, and the Connect button.

Its job:
  • HTML — HyperText Markup Language: defines the page’s structure: headings, text, buttons, menus, and form fields.
  • CSS — Cascading Style Sheets: defines how that structure looks: colors, fonts, spacing, and the layout of pages and menus.
  • JavaScript: makes the page interactive — opening a submenu, noticing a click, sending a request, and updating the screen when the answer returns.

Example: React

A JavaScript library often used to build reusable parts of the screen, such as a profile card or dropdown, and keep them updated as data changes.

02 · The features and rules · back-end

Service layer

The service layer is the code — the programming — that defines what the software does and how its features work: connecting with a person, sending a message, saving a setting, or deciding who can access a document. It runs on the company’s servers. DuckBuddy’s services are written in Python.

Its job:
  • HTTPS — HyperText Transfer Protocol Secure: the rules for sending requests and responses between your browser and the company’s servers over an encrypted connection. It protects information while it travels; it does not decide who is allowed to see it.
  • Run the feature’s rules: check who is signed in, what their access level allows, and whether the request is valid — for example, whether they may change this account setting.
  • Work with stored information: ask the database to find or save the right records, then send an answer back so the front-end can display the result.

Example: Saving an account setting

When Anna turns email notifications off and clicks Save, the browser sends an HTTPS request. The service layer checks that Anna is signed in and may change that account, validates the setting, asks the database to save “off” for her email notifications, and sends a confirmation back. The front-end then displays the saved setting.

03 · The storage · also back-end

Database

The database remembers information after you close the tab: user settings, which account options are turned on or off, each user’s access level, profiles, and saved dropdown choices. It also stores records about uploaded files and documents, such as their names, owners, and storage locations; the files themselves often live in separate file storage. DuckBuddy uses several databases, including PostgreSQL and Redis.

Its job:
  • Create: add information — such as a new account or a record for an uploaded document.
  • Read: retrieve information — such as your saved settings or the choices for a dropdown.
  • Update: change information — such as turning a setting off or changing a user’s access level.
  • Delete: remove information — such as an obsolete saved option. Together, these four operations are called CRUD.

Database versus Database Schema

  • Database: the organized store that holds the schema and the actual saved information, and lets the service layer retrieve or change it.
  • Database schema: the blueprint for organizing stored information: tables or collections, fields (columns in a table), data types, relationships, and rules. For example, a settings table might define user ID, setting name, and an on/off value.
  • Data: the actual values stored within that blueprint — for example, Anna’s user ID, “email notifications,” and “off,” or the saved list of department options in a dropdown.

Example: Menus and submenus

When a company stores menu definitions in a database, the schema defines fields such as label and parent menu ID. The data supplies actual values — for example, “Account” as a menu and “Notifications” as its submenu. The front-end still draws their look and layout. Not every menu or dropdown is database-driven.

Where they sit

This is the part most people never picture. Only one layer runs on your computer. The other two stay on the company’s machines in the cloud.

On your computer or phone
Your browser
DuckBuddy · Anna’s profile
Connect · Chat · Search

Front-end runs here

When you open the site, the browser downloads the front-end files and runs them on your machine. That is why the page can feel instant — the buttons are already on your computer.

The files themselves are stored on company servers. They only run once they arrive in your browser.

Company cloud — not your laptop
DuckBuddy servers · Python Always on. Waiting for HTTPS from browsers, phones, and other servers.

Service layer lives here

This code never runs on your computer. You cannot open it, edit it, or see it. It sits in a data center (AWS, Google Cloud, Azure, and others) so it can serve millions of people at once.

If you close your laptop, the service is still there, still answering other people’s requests.

Also company cloud — locked away
PostgreSQL · profilesAnna’s name, job, location
Redis · “online now”Fast memory, not a file on your phone
Neo4j · who knows whomThe friend graph

Database lives here

Your browser is not allowed to talk to the database. Only the service layer is. That is how a company keeps other people’s data off your laptop.

One sentence: the front-end runs on the user’s computer after it is downloaded; the service and the database stay on the company’s servers. HTTPS crosses the public internet. The service-to-database call usually stays on a private network you never see.

Front-end

Stored: company servers (so you can download it).
Runs: your browser or phone app.
You: can view the page source. You cannot see other people’s private data from here alone.

Service layer

Stored and runs: company cloud.
You: never run this code. Your computer only sends it an HTTPS request and waits for a packaged answer.

Database

Stored and runs: company cloud (sometimes a database company’s cloud).
You: never connect to it. If the front-end could, anyone could Read or Delete everyone else’s rows.

How they get programmed

Different people, different languages, different “done.” A feature is not finished until all three have been written, tested, and deployed.

Written for the browser

Front-end

Ex:HTMLCSSJavaScriptReactAngular

HTML is the structure (a button, a list). CSS is the look (color, layout). JavaScript (and React) is the behavior — “when they click Connect, send this request.”

A front-end developer builds screens. They do not usually write the database tables. They call the service and paint whatever comes back.

Shipped as: files your browser downloads. Change the button label? That is a front-end change.
Written for the server

Service layer

Ex:PythonNode.jsJava.NETPHPRuby

A back-end developer writes a program that listens for HTTPS (GET, POST, and friends), checks login and permissions, then sends CRUD to the database.

This is also where “you may only see Texas contacts” or “do not connect twice” lives. Rules do not belong in the browser — anyone could skip them.

Shipped as: a service that stays running in the cloud. Change who is allowed to Connect? That is a service change.
Designed, then queried

Database

Ex:SQLPostgreSQLMySQLMongoDBRedis

Someone designs tables or collections (a row per person, a row per like). The service then talks in CRUD: Create a connection, Read Texas contacts, Update a like count, Delete almost never in banking.

SQL is the common language for “find / add / change / remove.” MongoDB and Redis use different commands, but the four jobs are the same.

Shipped as: a running database plus a schema (what columns exist). Add “hobbies” to a profile? That often needs a new column — a database change.

A “small” feature can touch all three: a new field on the screen (front-end), a new check in the service, and a new column in the database. That is the whole point of this explorer.

How they interact

One click. Seven steps. Nothing on your screen changes until the packaged answer comes back.

1

You click something

Connect, Add to cart, filter Texas, Play — the front-end on your computer notices.

Runs on your computer
2

The front-end sends HTTPS

Your browser packages a request (GET /contacts?state=TX or POST /connections) and sends it across the public internet to the company’s servers.

Leaves your computer
3

The service layer receives it

Ex: Python, Node, Java, or similar — still on the company cloud — checks: are you logged in? Are you allowed? Is this request complete?

Company servers
4

The service asks the database

A CRUD command goes over the company’s private network. Your laptop is not on this call. Read contacts in Texas, or Create a cart line.

Private network · not the public internet
5

The database answers the service

Rows come back to the service only — Jordan in Austin, Priya in Dallas. The database does not talk to your browser.

Still on company servers
6

The service packages an HTTPS response

It turns those rows into JSON (or a page) and sends that package back across the public internet to your browser.

Back across the internet
7

The front-end updates the screen

React (or the page) paints the new list, fills the heart, or says Connected. You still never touched the database yourself.

Runs on your computer again

Mobile apps follow the same seven steps. Ex: the front-end is Swift, Kotlin, or React Native on the phone instead of a browser — but the HTTPS still goes to the same service and the same database.

Why all three are needed

If one layer is missing, the product breaks in a specific way. That is why “just change a few lines of live code” is the wrong picture.

Without a front-end

Nothing to click

The service and database could be perfect, and you would still have no screen. No dropdown, no cart button, no profile. People do not talk to Python or SQL. They talk to a UI.

Without a service layer

Unsafe, and no rules

If the browser talked to the database directly, anyone could Read or Delete other people’s rows. The service is the lock on the door: login, permissions, “don’t connect twice,” “only Texas if you’re allowed.”

Without a database

Nothing remembered

Close the tab, and Anna’s profile, your cart, and yesterday’s like would vanish. A database is why the facts are still there tomorrow — on the company’s machines, not on yours.

Front-end without the other two is a poster: it can look like DuckBuddy and never save a connection. All three together are the product.

Where to Go Next

Keep going in order, or jump to any page from the top menu.