Inspiration

During online conversations, many platforms either compromise privacy or lack safety features. Apps like Omegle were fun but unsafe. We wanted to reimagine random chatting — making it secure, anonymous, and engaging, while also adding modern features like avatars, reporting, and real-time translation. Our goal: “Connect people randomly, but responsibly.”

What it does

P(user gets matched in k tries)=1−(1−n1​)k

How we built it

Frontend ( React + TailwindCSS) Generated a clean, responsive UI with React. Added chat UI components: message bubbles, avatars, typing indicator, “Next” button. Backend (Supabase + Node.js) Authentication: Guest/username login via Supabase Auth. Matchmaking: Used a waiting_users table → when two users are queued, they’re paired into a rooms table. Chat Messages: Stored in a messages table, streamed with Supabase Realtime. Moderation: Profanity filter applied before saving/broadcasting messages. Safety: Added reports table for user reporting. Extra Features Random avatar + nickname generator. Typing indicators via realtime events. “Next” button → instantly leave a chat and find someone new

Challenges we ran into

Matchmaking Logic: Ensuring no deadlocks (e.g., when many users join/leave quickly). Realtime Updates: Keeping chat messages and typing indicators synced without delay. Database Race Conditions: Handling two users trying to join a room at the same millisecond.

Accomplishments that we're proud of

Built a working random chat app with authentication, matchmaking, real-time messaging, and moderation — all within hackathon time limits. Designed a safe alternative to Omegle, with features that make conversations fun but also secure.

What we learned

Supabase as a Backend – how to use Supabase Auth, Realtime, and Database to replace a traditional backend. Matchmaking Logic – maintaining a queue of users and pairing them randomly felt like solving a probability puzzle. Real-Time Communication – setting up live chat with typing indicators taught us the importance of event-driven systems.

What's next for random-connect

AI-powered toxic message detection for better safety. Multi-language auto-translation of messages in real-time.

Built With

  • ai-assisted
  • database
  • frontend:-react
  • generation
  • github
  • loveable-(for-ui-generation)-backend-/-realtime:-supabase-(auth
  • markdown-(for-documentation)
  • npm
  • profanity-filter-(custom)-hosting-&-deployment:-vercel-(frontend)
  • realtime)-database:-supabase-postgres-apis-&-services:-supabase-realtime-api
  • supabase-(backend/db)-other-tools:-node.js
  • tailwindcss
  • typescript
Share this project:

Updates