I'm 19, studying software engineering in Karachi, Pakistan and a
month ago I decided to build something real. Not a todo app. Not a
clone. Something I could actually charge money for.
That thing is FeedbackLoop | an AI powered web and code reviewer.
You paste a URL or code snippet and get back a scored list of issues
with exact fixes for each one. SEO, performance, security,
accessibility and UX all in one report.
Here's how the month actually went.
The idea
I got tired of finishing a website and having no idea if it was
actually good. Existing tools either cost a fortune or gave you
a score with zero explanation of what to actually do about it.
I wanted something that said "here's what's broken and here's
exactly how to fix it." Simple as that.
The build
Stack I went with:
- Next.js 14 with App Router
- Supabase for database and auth
- Groq API for AI analysis
- Browserless for full JS rendering
- Tailwind CSS
- Deployed on Vercel
Sounds straightforward. It was not.
The part that nearly broke me
The API integration was genuinely painful. I started with Gemini
as my AI provider and spent days getting everything connected and
working. Then it broke. Just stopped working the way I needed it to.
Had to rip it out and switch everything over to Groq. That was not
a fun afternoon.
But honestly it was worth it | Groq is faster and has a generous
free tier which matters a lot when you have zero users and zero
revenue.
What I actually built
By the end of the month FeedbackLoop had:
- URL and code review with severity scored issues
- Individual scores for SEO, Performance, Security, Accessibility and UX
- Full JavaScript rendering so React and Next.js apps get properly analyzed
- WordPress specific fix guides in plain English for non-technical users
- Plugin recommendations for WordPress sites
- Downloadable score card to share your results
- 3 plan tiers | Free, Developer ($9/mo), Pro ($19/mo)
- Early bird deal | first 100 users get Pro free for 3 months
Why SaaS and not just another portfolio project
Honestly I just wanted to do something different. Every software
engineering student has the same portfolio | todo app, weather app,
maybe a clone of something. I wanted to build something that could
actually make money and solve a real problem.
Even if FeedbackLoop fails completely I learned more in this one
month than in any classroom. Payments, auth, APIs, rate limiting,
database design, deployment, marketing | all of it.
Where I am now
Just launched. Zero paying users. Zero free users if I'm being
completely honest.
But it's live, it works, and I'm proud of it.
If you want to try it | no signup needed, there's a free preview
right on the homepage.
feedbackloopai.vercel.app
And if you're a student thinking about building something real |
just start. It doesn't have to be perfect. Mine definitely isn't.
But it exists and that's more than most people can say.
Top comments (0)