RapidGig

A video-first job platform connecting freelancers and recruiters with AI-powered video generation.

Live Demo

https://rapid-gigs.vercel.app/

Features

• AI Video Generation — 30-second promotional videos stitched from cinematic prompts using multi-provider failover (Pollinations, Replicate, fal.ai)
• Video Shorts Feed — TikTok-style vertical video feed for job introductions
• User Authentication — Email/password + Google OAuth login
• Job Listings — Create, browse, filter, and apply to jobs
• Applications — Students apply with resumes and video introductions; recruiters review and manage applicant status
• Real-time Messaging — Built-in chat between students and recruiters
• Role-based Access — Separate dashboards for students and recruiters
• Cloud Storage — File uploads stored on Cloudinary (persistent across deploys)

Tech Stack

Architecture

┌─────────────┐     HTTPS     ┌─────────────────┐
│  Vercel     │ ────────────▶ │     Render      │
│  (React)    │               │  (Express API)  │
│             │ ◀──────────── │                 │
└─────────────┘   JSON/API    └────────┬────────┘
                                       │
                              ┌────────┴────────┐
                              │                 │
                       ┌──────────────┐  ┌──────────────┐
                       │  MongoDB     │  │  Cloudinary  │
                       │  Atlas       │  │  (files)     │
                       └──────────────┘  └──────────────┘

Getting Started

Prerequisites

• Node.js 22+
• MongoDB (local or Atlas URI)

Quick Start

git clone https://github.com/Kiranpjk/RapidGigs.git
cd RapidGigs
npm run dev

This starts:

• Backend API on http://localhost:3001
• Frontend app on http://localhost:5173

Manual Setup

Backend:

cd backend
npm install
# Create .env (see Configuration below)
npm run dev

Frontend:

cd frontend
npm install
npm run dev

Run tests:

npm run dev:backend   # backend only
npm run dev:frontend  # frontend only

Configuration

Create backend/.env:

PORT=3001
NODE_ENV=development
MONGO_URI=mongodb://localhost:27017/rapidgig
JWT_SECRET=your_secret_here
CORS_ORIGIN=http://localhost:5173

# Google OAuth
GOOGLE_CLIENT_ID=your_client_id.apps.googleusercontent.com
GOOGLE_CLIENT_SECRET=your_client_secret

# Cloudinary (free tier: 25GB)
CLOUDINARY_CLOUD_NAME=your_cloud_name
CLOUDINARY_API_KEY=your_api_key
CLOUDINARY_API_SECRET=your_api_secret

# AI Services (free tiers)
GROQ_API_KEY=your_groq_key
OPENROUTER_API_KEY=your_openrouter_key
FAL_KEY=your_fal_key
REPLICATE_API_TOKEN=your_replicate_token
POLLINATIONS_KEY=your_pollinations_key

Create frontend/.env:

VITE_API_BASE=http://localhost:3001/api
VITE_GOOGLE_CLIENT_ID=your_client_id.apps.googleusercontent.com

API Endpoints

AI Video Pipeline

1. Prompt Enhancement — Job description → cinematic video prompt via OpenRouter → Groq → Ollama
2. Clip Generation — 3×10s segments via multi-provider failover:
   • Pollinations (free)
   • Replicate (free tier)
   • Hugging Face Router
   • fal.ai (pay-per-use)
3. Image-to-Video Stitching — Last frame of each clip feeds the next for continuity
4. Concatenation — ffmpeg stitches clips into one 30s video

Deployment

• Frontend: Vercel — auto-deploys on push to main
• Backend: Render — build: npm install && npm run build, start: npm start
• Database: MongoDB Atlas (free tier, 512MB)
• Storage: Cloudinary (free tier, 25GB)