All micro-internships
Micro-internship brief ≈15h across 2 daysbuilder

Ship an app that mashes a real API with AI

Most real products are plumbing: take data from one place, add intelligence, show it usefully. Build and deploy an app that combines a real public API with an LLM — live data in, AI-processed insight out.

AI Application DevelopmentAnalytical ReasoningWeb DevelopmentAPI Integration

Posted by The AI Internships

The work brief

  1. 01Pick a free public API with live data: Open-Meteo (weather), OpenAQ (air), Frankfurter (currency), a sports score API, or your country’s open-data portal.
  2. 02Design the mashup: what does the AI ADD to the raw data? (e.g. weather → "should I water the plants today?" advice in a language used in your community; currency → plain-language trend explanation).
  3. 03Build and deploy it with AI-assisted code (free host). Handle the ugly parts: API failures, rate limits, weird data.
  4. 04Push code to a public repo and document one case where the AI’s output was wrong or unhelpful, and how you constrained it.

What you’ll produce

4 deliverables

Submission standard

Submit the live URL, repo, what the AI adds beyond the raw data, and the failure case you handled. We’ll use the app. Your repo must show the mashup built and hardened across several commits, not one upload.

  • Live URL of your deployed app

    Public linkRequired

  • Public GitHub repo

    Public GitHub repo — commit history should show the API wiring, deploy, and error-handling fixes across multiple commits, not a single paste. A one-commit dump fails the authenticity gate.

    Public linkRequired

  • The mashup: which API, what the AI adds, why that’s useful

    API means application programming interface. Name the live data source, what the AI changes or explains, and why the combination is useful.

    Written responseRequired

  • One case where the AI output was wrong/unhelpful, and how you constrained it

    Written responseRequired

You’ll complete these inside your private workspace.

How your work is evaluated

The passing benchmark is 70/100.

Live mashup

29%

The app works live: real API data flows in, AI output comes out.

AI adds value

29%

The AI layer genuinely transforms the data — not a chatbot bolted on.

Engineering honesty

14%

Failure handling is real and documented; code is in the repo.

Live + built over commits

29%

If the live URL is dead (no real API data flowing), or the repo is a single copy-paste commit, this fails regardless of how the mashup is described.

How we grade your AI usage

30% of your score

Using AI is the point — it’s the skill this certificate proves. You’ll answer three short questions about how you used it: what you asked, what was wrong with its first answer, and what you changed. Specific, honest answers score high. “I pasted the brief and submitted the answer” scores near zero.