This is a submission for the World's Largest Hackathon Writing Challenge: After the Hack.
🧠The Problem: Connection in the Age of Disconnection
Before the hackathon even began, I was wrestling with a question that felt deeply personal: Why is it so hard to meet people in real life, even though we’re more digitally connected than ever? I wasn’t interested in building another dating app or generic social media feed. I wanted something that made it easier for people to say, “I’m going to this park,” or “I want to try this coffee shop,” and actually find others who want to join.
So I built Friender.
Friender is a social app for people who are tired of swiping, ghosting, and endless small talk. It helps you meet others through shared activities and interests. No bios, no AI-generated icebreakers, no pressure to impress. Just real-world intentions and people who are down to join.
But here’s what makes it truly different — Friender doesn’t show you a public feed of random events. Instead, it matches you with people first based on deep signals like behavior, preferences, and intent. Then, and only then, it shows you the things your matches are planning to do. This ensures that every activity you see is already socially relevant and meaningful to you.
🚧 The Build: From Idea to Working Prototype
I started by asking myself: What would make me actually use a social app like this?
Key Features I Designed:
- People-first, activity-later: Friender matches you with compatible people before showing you their activities. No overwhelming public feed of strangers.
- Intent-based matching: Matches are determined by shared lifestyle signals, not profile photos or bios.
- Data-driven enrichment: I used Spotify and OpenAI to analyze user taste and generate personality summaries. This creates a light and expressive onboarding process without friction.
- No swiping, ever: That mechanic is built for quick judgment. Friender removes it entirely and focuses on shared interest and vibe.
Technologies Used:
- Frontend: React, TailwindCSS
- Backend: Python, Flask
- Authentication: Supabase Auth
- APIs: Spotify Web API for music taste analysis, OpenAI for smart profile enrichment and optional clarifying conversation prompts
- Matching Algorithm: Lightweight similarity scoring based on music taste, behavioral data, time preference, and shared activity types
- Hosting: Render (backend), Vercel (frontend)
- Database: PostgreSQL (via Supabase)
A major challenge was personalizing the experience without making users fill out long surveys. That’s where social integrations came in. Spotify tells us what people listen to. Google Play Games shows behavior patterns. OpenAI helps create friendly, natural-sounding summaries that keep things human. And all this happens without needing users to spend more than 30 seconds onboarding.
đź’ˇ What I Learned: Technical and Personal Growth
This project changed me, in more ways than I expected.
Technical Skills Gained:
- I gained deep hands-on experience with OAuth integrations and Spotify’s API, including user consent flows and token management.
- I experimented with OpenAI's API for user enrichment, learning how to use AI in a human-focused way rather than just for novelty.
- I became more fluent with Supabase, using its tools for real-time sync, authentication, and data querying.
- I refined my UX skills to focus on clarity, comfort, and simplicity, avoiding the growth-hack patterns of most social apps.
- I practiced building and testing fast, within real-world constraints, while maintaining code readability and modularity.
Personal Transformation:
- I came into the hackathon thinking I was just here to ship a prototype. But midway through, I realized I actually cared about the problem.
- I stopped focusing on whether it would win. I started thinking about whether it was worth building.
- I became more comfortable pitching real ideas, even if they break conventional patterns.
- I learned that meaningful work doesn’t have to be polished — it just has to be honest and intentional.
🚀 What’s Next for Friender
Friender is no longer just a hackathon project. I want to keep pushing it forward and see what real people think about it.
Future Plans:
- Improve the matching algorithm with more social graph enrichment using other OAuth sources like Google Play Games, YouTube, or Reddit.
- Add optional personality calibration questions for those who don’t use Spotify.
- Design a privacy-first backend that gives users full control over what’s shared and how it’s used.
- Launch a public-facing landing page to gauge early interest and open up email waitlists.
This app was built in just 10 days. I heard about the hackathon late, and development began fast. But even in that short time, Friender grew into something that I want to see come alive — not because it’s perfect, but because it’s needed.
❤️ Final Thoughts
This hackathon reminded me that building something with purpose will always be more valuable than building for clout.
Friender is not a viral stunt. It’s a small attempt to make tech more human. It helps people say, "I want to do this — who wants to join?" and makes that feel natural, safe, and fun.
Thanks to the organizers for making this space to create and reflect. And thank you to everyone who’s ever questioned why we need a better way to meet people.
If this resonated with you, I’d love to hear your thoughts or collaborate. Let’s keep creating for the people, not just the feed.
Let’s build a world where friendship starts with doing things together.
Top comments (0)