Learning New Things Is Hard: But Here’s Something Pretty Cool I Learned Anyway

To be completely honest with you, learning new things — no matter how much you want to — is hard. I personally blame TikTok and Instagram for the tragic state of my attention span.

But while my brain is still buffering on “new things,” let me talk about something I did manage to learn (and actually remember).
Something I now bump into almost every single day as a Graduate AI/Automation Engineer:

API: Application Programming Interface

Textbooks will give you a long, delicate, clinically precise definition — and those are great, truly.
But here’s how I understood it in a way that stuck:

APIs are structured ways for one system to talk to another.

Somewhere out there, people or organisations are sitting on data in their databases. This data is theirs — private, structured, protected.
You can’t just stroll in and grab it.

But sometimes they want (or need) to share parts of it — either because it’s useful, profitable, necessary, or required by how the modern web works.

So they create a clean, safe, controlled door for others to access selected pieces of their data.

That door is the API.

So What Does an API Actually Look Like?

An API usually exposes endpoints, which are basically URLs that point to specific resources.

For example:

https://api.example.com/v1/users
https://api.weather.com/v3/forecast/daily
https://mycoolstartup.com/api/products/42

Each endpoint represents something you can interact with — a list of users, weather data, a specific product, a record in a database, etc.

You interact with it using HTTP methods. The main four are:

GET → “May I please have this data?”
POST → “Here’s some new data to store.”
PUT → “Update this existing thing with new info.”
DELETE → “Remove this thing. Yeet it.”

But those are just the basics.

The Technical Sauce Behind APIs

Let’s talk about what’s actually happening under the hood.

Structured Responses: JSON

Most modern APIs respond using JSON, a clean, predictable text format that looks like this:

json
{
  "user_id": 67,
  "name": "Elo",
  "hobby": "breaking servers"
}

Your machine receives this, parses it, and now you’ve got structured data to work with.

Parameters: How You Ask for Specific Stuff

Endpoints often accept parameters:

Query Parameters (for filtering, sorting, searching)

https://api.example.com/v1/products?category=shoes&limit=20

Path Parameters (to fetch something specific)

https://api.example.com/v1/products/1234

Body Parameters (for POST/PUT)

json
{
  "name": "AI-powered Toaster",
  "price": 7999
}

Headers: The API Bouncers

Headers are like ID badges.
They tell the server things like:

What format you expect (application/json)
Who you are (Authorization: Bearer )
What app or script is calling

No headers = no entry in many APIs.

Authentication: The Gatekeeping

Most useful APIs need some form of authentication:

API Keys
OAuth Tokens
JWT (JSON Web Tokens)
HMAC signatures (more secure, used in banking/fintech APIs)

This is how the data owners ensure you’re allowed to access what you’re asking for.

---

Where Things Get Really Interesting: Rate Limits & Errors

APIs don’t let you go wild.

Rate Limits

You often get something like:

100 requests per minute
10,000 calls per day
or “Stop spamming me pls”

Break the limit and you get hit with:

429 Too Many Requests

Error Codes

Some classics include:

400 → You messed up the request
401 → You’re not authenticated
403 → You’re authenticated, but still not allowed
404 → Endpoint not found
500 → Server had a meltdown
503 → Server said “I need a nap”

These codes become your spiritual companions as a developer.

---

Why APIs Are the Unsung Heroes of Modern Tech

Almost everything you use daily involves APIs:

Google Maps loading directions
Uber figuring out nearby drivers
Your bank checking your balance
Instagram loading posts (ironically the reason I can't focus…)
AI systems pulling data to make decisions

And for me?

A massive chunk of my job is essentially:

1. Calling an endpoint
2. Authenticating properly
3. Fetching the data
4. Cleaning it
5. Feeding it into an AI or automation pipeline

Sometimes that pipeline triggers Slack alerts.
Sometimes it updates CRM systems.
Sometimes it nudges internal dashboards.
Sometimes it powers an entire AI model.

APIs are the pipes, and I’m basically the plumber.

A glamorous plumber.

---

The Beautiful Part: You Don’t Need to Know Everything to Start

I used to think APIs were some mystical, highly technical thing that only “real developers” understood.

Then I realised:

At their core, APIs are predictable, structured conversations between systems.

And if you can read a JSON object, construct a URL, and send a GET request, you’re already halfway there.

Everything else is just layers you learn over time, attention span or not.