The Shree Swaminarayan Temple in Secaucus needed a platform for their 25th Anniversary celebration to handle event registration, admin workflows, display media content, share event schedules, and keep the community up to date on the latest information. Not just a brochure site, but a productionized product.
Why I Took This On
When I heard about the project, I was using Sonnet 4 to write love letters to my girlfriend (jk but actually..). I was fascinated by it. It finally made my girlfriend laugh.
I wanted to test its limits. I'm a backend and distributed systems engineer and I'd never built a production frontend, but curiosity won out, so I decided to partner with Sonnet 4 on this.
That decision ended up being one of the best I've made. It put me ahead of the curve on AI-driven development and got me so invested into the ever-developing world of agentic coding that my entire X feed is all about the latest Ralph Wiggum plugin.
Built Quickly, but Learned Faster
A lot of people use AI as a crutch to avoid work they don't want to do. I was sometimes guilty of that too (but who actually likes writing unit tests?). But what I found is that the real unlock was treating AI as a thought partner and educator. I'd ask how it would implement things, push back on suggestions, walk through system design options, and at the end of a 30 minute back and forth, I'd have a deep understanding of concepts I didn't know existed an hour earlier.
I unlocked the ability to learn and build at the same time.
I applied my engineering experience (thinking through tradeoffs, infrastructure decisions, system design) and paired that with Claude's speed to build a production-grade product. I wasn't handing over the reins and saying "just do it." I learned why SSG > SSR > CSR, how CDNs affect Core Web Vitals, how to build mobile-first, how Cloudflare R2's image optimization pipeline beats paying for Vercel Pro, and how Supabase works as a backend service. All things I didn't know before I started. All things I know well now.
Working with Real Customers
I didn't just learn from AI on this project. I learned from real humans too. Working directly with end users of my software, I saw firsthand that quick iteration earned trust.
When organizers asked for community service submission forms and progress tracking against their benchmarks, I shipped it fast. Event organizers needed visibility into registrations, community service submissions, and key metrics. So I built an admin portal with Google OAuth for access control, data views with filtering, and keyset pagination so the tables stayed fast as data grew.
That opened the door to more responsibility and more impactful work.
The Takeaway
This project was an eye-opener. AI helps you move faster, but the real value is the education it provides on entirely new concepts. When you combine that with the ability to think through problems and anticipate issues before they happen, you get better outcomes than either side could produce alone.
Check out the live website to see the platform in action, or browse the GitHub repo for the full codebase and README. If you have questions or feedback, I'd love to hear from you. Reach out anytime.