Ship an MVP for a problem you can see from your house
The capstone: find a real problem in your own college, street, or community, ship an AI-built, human-finished MVP that addresses it, and put it in front of 3 real potential users.
Posted by The AI Internships
The work brief
- 01Find a problem you’ve personally observed (hostel complaints go nowhere, no one knows the local bus timings, tuition teachers can’t manage bookings).
- 02Talk to 2-3 affected people BEFORE building — confirm it’s real and note their words.
- 03Build and deploy the smallest thing that helps — AI writes the first draft, you fix and finish it — live on a public URL with the code on GitHub.
- 04Show it to 3 real potential users and record their honest reactions — including the negative ones.
What you’ll produce
Submission standard
Submit the problem evidence, live URL, repo, and the 3 user reactions. All-positive feedback is a red flag — real users always want something changed. Include your AI workflow: tools, best prompts, and what you changed from the AI’s output. Your repo must show the MVP built across several commits, not a single upload.
The problem, who has it, and what the affected people said before you built
Written responseRequiredLive URL of your deployed MVP
Public linkRequiredPublic GitHub repo
Public GitHub repo — commit history must span the build (multiple dated commits as you drafted with AI and fixed by hand), not a single paste. A one-commit dump fails the authenticity gate.
Public linkRequired3 real user reactions — quotes, what confused them, what they’d change
Written responseRequired
You’ll complete these inside your private workspace.
How your work is evaluated
The passing benchmark is 70/100.
Real problem
29%The problem is specific, local, and validated with real people before building.
Shipped MVP
29%The live URL works and genuinely addresses the stated problem.
Real user feedback
14%Reactions read authentic, with specifics and criticism.
Real validation + real build trail
29%If the live URL is dead, the repo is a single dump commit, or the feedback is all-positive/generic with no pre-build validation, this fails regardless of prose polish.
How we grade your AI usage
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.