job-eligibility

Job Eligibility Checker

A full-stack web app that evaluates how well your profile matches a target job role. You fill in your education, skills, experience, and certifications — and it gives you a detailed score breakdown, skill gap analysis, and actionable recommendations to improve your chances.

Built with a React + Vite frontend and a Python FastAPI backend running a trained scikit-learn model.

Live: job.gaury.dev · API: job-api.gaury.dev

---

What it does

• Scores your profile across 4 dimensions — skills, education, experience, and academics
• Classifies you into tiers: Highly Eligible, Eligible, Partially Eligible, or Foundational / Action Required
• Shows which skills you match, which ones you're missing, and any bonus skills you bring
• Generates prioritized recommendations (what to learn, what to build, what to certify)
• Supports 6 job roles: Python Developer, Data Analyst, ML Engineer, Data Scientist, Web Developer, Software Developer

Tech stack

Frontend

• React 19, TypeScript, Vite
• Tailwind CSS v4
• Framer Motion (animations)
• Lucide React (icons)

Backend

• Python 3.12, FastAPI, Uvicorn
• scikit-learn (trained classification model)
• Pydantic v2 (request/response validation)
• Pandas, NumPy, Joblib

Hosting

• Frontend on Vercel
• Backend on Render

---

Project structure

├── src/                          # React frontend
│   ├── components/               # UI components (Hero, Form, Results, etc.)
│   ├── services/                 # API client (eligibilityService.ts)
│   ├── data/                     # Job roles catalog
│   ├── types/                    # TypeScript interfaces
│   ├── context/                  # Theme provider (dark/light mode)
│   └── utils/                    # Animation helpers
│
├── backend/                      # Python FastAPI backend
│   ├── main.py                   # App entry point, CORS, routes
│   ├── ml/                       # ML model, predictor, preprocessing
│   ├── models/                   # Trained .joblib model file
│   ├── services/                 # Eligibility engine, skill matcher, recommendations
│   ├── schemas/                  # Pydantic request/response models
│   ├── data/                     # Training dataset
│   └── tests/                    # Pytest test suite
│
├── index.html                    # Vite HTML entry
├── vite.config.ts
├── package.json
├── render.yaml                   # Render deployment config
└── Procfile

---

Getting started

Prerequisites

• Node.js 18+
• Python 3.10+

Frontend

npm install
npm run dev

Opens at http://localhost:3000. By default it'll call the production backend API — you can override this with a .env.local file:

VITE_PYTHON_API_URL=http://localhost:8000

Backend

cd backend
pip install -r requirements.txt
cd ..
uvicorn backend.main:app --reload --port 8000

API docs will be at http://localhost:8000/docs (Swagger UI).

Running tests

cd backend
pytest

Google OAuth Setup

The app supports Google Sign-In via Supabase Auth. No frontend environment variables are required — all OAuth configuration is done in the Supabase and Google Cloud dashboards.

1. Google Cloud Console

1. Go to Google Cloud Console → Credentials
2. Create an OAuth 2.0 Client ID (Web application)
3. Add Authorized JavaScript origins:
   • https://job.gaury.dev
   • http://localhost:3000
4. Add Authorized redirect URIs:
   • https://zgtrtrxearlhoqmyyhrp.supabase.co/auth/v1/callback
5. Copy the Client ID and Client Secret

2. Supabase Dashboard

1. Go to Authentication → Providers → Google
2. Enable the Google provider
3. Paste the Client ID and Client Secret from Google Cloud Console
4. Under Authentication → URL Configuration, add these to Redirect URLs:
   • https://job.gaury.dev
   • http://localhost:3000

3. How it works

• The frontend calls supabase.auth.signInWithOAuth({ provider: 'google' }) with redirectTo set to window.location.origin
• Supabase redirects the user to Google's consent screen, then back to your app
• The existing onAuthStateChange listener picks up the new session automatically
• The database trigger handle_new_user() auto-creates a profile row for new Google users
• Existing RLS policies work identically — auth.uid() is the same for OAuth and email users

---

API

The backend exposes a single main endpoint:

POST /api/v1/evaluate-eligibility

Request body:

{
  "fullName": "Jane Doe",
  "educationLevel": "Bachelor's Degree",
  "branch": "Computer Science",
  "cgpa": "8.5",
  "technicalSkills": ["Python", "FastAPI", "Docker", "PostgreSQL"],
  "yearsOfExperience": "2",
  "certifications": ["AWS Cloud Practitioner"],
  "targetRole": "python-developer"
}

Response includes overall score, tier classification, per-dimension scores, matched/missing/bonus skills, and prioritized action recommendations.

Other endpoints:

• GET / — service info
• GET /health — health check + model status

---

Deployment

The repo is set up for split deployment:

• Frontend → Vercel — auto-detected as a Vite project, builds to dist/
• Backend → Render — uses render.yaml for config, runs uvicorn backend.main:app

The key env vars you need:

Where	Variable	Value
Vercel	VITE_SUPABASE_URL	Your Supabase project URL
Vercel	VITE_SUPABASE_PUBLISHABLE_KEY	Your Supabase anon/public key
Vercel	VITE_PYTHON_API_URL	Your Render/backend URL
Render	PYTHON_VERSION	3.12.0
Render	CORS_ORIGINS	Your Vercel/frontend URL
Supabase	Google OAuth Client ID \u0026 Secret	Configured in Dashboard → Auth → Providers → Google

---

License

MIT