⚡ EmploySure — AI-Powered Real-Time Job Search Agent

A free, zero-paywall, anti-scam job search aggregator that scrapes job listings from any URL, filters out garbage using AI, and presents a clean, unified dashboard.

No resume upsells. No AI cover letter filler. No dead links. Just real jobs.

---

🏗️ Tech Stack

Layer	Technology
Frontend	React 19 + TypeScript + Vite
Styling	Vanilla CSS (custom design system with glassmorphism & gradients)
Backend	Python 3.11+ + FastAPI
Scraping	crawl4ai (Playwright-based headless browser)
AI Primary	Google Gemini 2.5 Flash (free tier)
AI Fallback	Groq Llama 3.3 70B (free tier)
AI Offline	Ollama (local models, optional)
Database	SQLite (dev) → PostgreSQL (production)
Scheduling	APScheduler (async background jobs)

---

🚀 Quick Start

Prerequisites

• Python 3.11+ — Download
• Node.js 18+ — Download
• At least one AI API key (free):
  • Google AI Studio → Get a Gemini API key
  • Groq Console → Get a Groq API key
  • Or install Ollama for local AI (no key needed)

1. Clone & Configure

# Navigate to the project
cd EmploySure

# Set up backend environment
cd backend
cp .env.example .env
# Edit .env and add your API key(s)

2. Start the Backend

cd backend

# Create virtual environment
python -m venv .venv

# Activate it
# Windows:
.venv\Scripts\activate
# macOS/Linux:
source .venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Setup crawl4ai (installs Playwright browsers)
crawl4ai-setup

# Start the server
python -m app.main

The backend will start at http://localhost:8000.

3. Start the Frontend

cd frontend

# Install dependencies
npm install

# Start dev server
npm run dev

The frontend will open at http://localhost:5173.

---

📋 Features

✅ Phase 1: Real-Time Multi-Source Scraping

• Add any URL — job boards, GitHub repos, company career pages
• Headless browser rendering (JavaScript-heavy pages supported)
• AI-powered anti-scam filtering (discards courses, MLM, resume services)

✅ Phase 2: Unified Dashboard

• Clean, spreadsheet-inspired data table
• One-click "Apply Direct ↗️" to the real application page
• Client-side instant filtering by title, experience, location
• Hide agency/recruiter postings with a single toggle
• Dark mode (default) + Light mode toggle

✅ Phase 3: Background Automation

• Hourly re-crawl of all active sources
• Automatic dead link detection and archival
• Database de-duplication (no duplicate job listings)

---

🔑 API Endpoints

Method	Endpoint	Description
POST	/api/sources	Add a new scraping target
GET	/api/sources	List all sources with status
DELETE	/api/sources/{id}	Remove a source
POST	/api/scrape	Trigger background scrape
GET	/api/jobs	List jobs (filterable, paginated)
GET	/api/jobs/stream	SSE for real-time updates
PATCH	/api/jobs/{id}/hide	Hide a listing
GET	/api/stats	Dashboard statistics
GET	/api/health	Health check

---

🎨 Design

• Dark Mode Default: Deep purple/indigo/violet gradient mesh background
• Glassmorphism: Frosted glass panels with backdrop-filter blur
• Micro-Animations: Button glow, row slide-in, pulsing status dots
• Light Mode: Clean black & white minimal theme
• Inter Font: Modern typography from Google Fonts
• Fully Responsive: Desktop, tablet, and mobile

---

📄 License

This project is free and open-source. Built to serve the community.