My First API Just Broke (And That Was the Point)

Titilola Olaitan — Fri, 17 Oct 2025 22:56:48 +0000

Day 1 of my internship. The task seemed simple: "Build an API endpoint that returns your profile and a cat fact."
How hard could it be? Spoiler: I learned more from this "simple" task than from watching 10 hours of tutorials.

The Challenge

Build a /me endpoint that:

Returns my profile info
Fetches random cat facts from an external API
Updates timestamp on every request
Doesn't crash when things go wrong

My Stack: Python + FastAPI

The "Aha!" Moments

External APIs Are Unreliable (And That's Normal) My first version looked like this: python@app.get("/me") async def get_profile(): response = await client.get("https://catfact.ninja/fact") cat_fact = response.json()["fact"] return {"fact": cat_fact, ...} What could go wrong?

API is down → My app crashes
API is slow → User waits forever
API returns weird data → App explodes

The fix: Timeouts + error handling + fallbacks
pythonasync def fetch_cat_fact(): try: async with httpx.AsyncClient(timeout=5.0) as client: response = await client.get(CAT_FACT_API_URL) return response.json()["fact"] except: return "Cat facts temporarily unavailable"

Lesson learned: Always plan for failure. Production apps need to be resilient, not just functional.

Environment Variables Aren't Just for Secrets I hardcoded my email in the code at first. Then I realized:

What if I want to change it?
What if someone else uses my code?
What about deployment?

Created .env for my actual data: bashUSER_EMAIL=john@example.com USER_NAME=John Doe Created .env.example for the template: bashUSER_EMAIL=your.email@example.com USER_NAME=Your Name
Lesson learned: Separate configuration from code. Future you will thank present you.

FastAPI's Auto-Docs Saved Me Hours

Hit localhost:8000/docs and boom.....interactive API documentation, automatically generated. I could test everything right in the browser without writing a single line of test code.

Lesson learned: Choose tools that give you superpowers for free.

The Final Result
Live API: [https://profile-api-production-f700.up.railway.app/me]
json{ "status": "success", "user": { "email": "john@example.com", "name": "John Doe", "stack": "Python/FastAPI" }, "timestamp": "2025-10-17T14:30:45.123456+00:00", "fact": "Cats sleep 70% of their lives." }
Every request fetches a new cat fact. Every timestamp is fresh. And when the cat API goes down? My API keeps working.

What This Task Actually Taught Me
Beyond the code:

Error handling > happy path — Production is messy
Async is powerful — Don't block your API
Never commit secrets — .gitignore is your friend
Documentation matters — README = first impression
Deploy early, deploy often — See your code in production

The unexpected lesson: A "simple" task reveals how much you don't know. And that's exactly the point of learning.

The Stack

Framework: FastAPI (because async + auto-docs = ❤️)
HTTP Client: httpx (for async requests)
Deployment: Railway (GitHub integration made it painless)
External API: Cat Facts API

Try It Yourself
GitHub: [https://github.com/Titilola-py/profile-api]
Live Demo: [https://profile-api-production-f700.up.railway.app/me]
Clone it. Break it. Fix it. Learn.

Questions? Built something similar? Drop a comment! I'd love to see what you created or hear about your first API experience.

Connect with me here:
GitHub: [https://github.com/Titilola-py] | LinkedIn: [https://www.linkedin.com/in/fatimah-olaitan-4bb6602a4/] | Twitter: [https://x.com/thetitilola]

DEV Community: Titilola Olaitan

My First API Just Broke (And That Was the Point)

The Challenge

My Stack: Python + FastAPI

The Stack