Web App with Chat GPT

 Modern app development has become dramatically simpler with AI-assisted coding.

Over the weekend, I experimented with an AI-driven workflow to go from idea to a working MVP. I wanted to understand how long will it take me to get to a working MVP on local. Answer - couple of hours (once product idea is finalized).

https://parentcircle-web.vercel.app/

My current workflow for building and validating a web app:
>> Requirements & data modeling (AI-assisted)
Using ChatGPT to generate and refine:
- Problem statement
- Feature specifications
- Database schema (tables, relations, constraints)
- API contract

>> Backend with Supabase
- Table creation and relational modeling
- Authentication setup with Row Level Security (RLS)
- Auto-generated REST APIs

>> Local development (AI-assisted)
Generate frontend components and service layers via ChatGPT
Integrate Supabase SDK for CRUD operations
Debug by feeding runtime errors and stack traces back into ChatGPT

>> Environment & configuration management
- Secure Supabase keys using environment variables
- Validate all flows locally

After local testing:
>> Deployment with Hostinger (pending – requires a plan)
- Build production assets
- Configure environment variables
- Attach custom domain

Tech Stack:
Frontend : Next.js (React) and tailwindCSS
Backend: Supabase
AI: OpenAI API

For an MVP, the productivity gains are significant.
This “vibecoding” approach shifts the mindset from writing code to designing systems and interaction flows.

Execution becomes faster; product thinking becomes the primary skill.

Where do I stand after just a few hours?
I discussed a random idea with ChatGPT, followed the steps above, and now have a fully working MVP running locally and connected to Supabase.

Honestly amazed at how much smoother the build process feels with ChatGPT. From architecture suggestions to debugging and integration, it feels like having a patient senior engineer available on demand.

I’ll be sharing more technical details and code examples on my blog soon.

Curious how others here are using AI in their development workflow.
What tools or stacks have you tried with vibecoding so far?

Flash card Quiz App

 This weekend, I turned that idea into something much more personal.

I built a small flashcard quiz app as a fun game for my daughter’s upcoming birthday.

The goal: learning + play + curiosity — all in one simple app.

What started as a “let’s try something fun” quickly became a fully working product:
🃏 Interactive flashcard game
🔄 Smooth card flip animations
📚 10 topics with Easy / Medium / Hard levels
🎯 One-word answers for quick learning
🎨 Clean, mobile-friendly UI
🚀 Deployed live on Vercel

Fun dev moment:
I initially integrated an OpenAI API call to generate questions dynamically…
and then quickly removed it to save costs 😅
Turns out, a well-structured local quiz bank works perfectly for a birthday game.

Tech stack used:
- Next.js (App Router)
- React + TypeScript
- Tailwind CSS
- Simple in-memory quiz bank (JSON)
- Vercel for deployment

Big reflection:
With today’s tools (and AI-assisted coding), it’s possible to go from an idea → UI → logic → deployment incredibly fast.

But more importantly, it reminded me that building small, joyful projects can be just as meaningful as building “serious” ones — especially when they’re for family.

Want to use the tool?
https://flashly-app.vercel.app/

Do give me your feedback. Looking forward to it.