Job Fraud Detection System - Anveshan Hackathon 2025

Team: Numeric Nomads

Video Link : google-drive link

Members:

• Bhawesh Agrawal
• Priyanka Singh (Team Lead)

---

📋 Project Overview

A Job Fraud Detection System that allows users to check if a job posting is likely fraudulent.

Tech Stack Used:

• Backend: FastAPI, Scikit-learn, PyTorch
• Frontend: Next.js (React), TailwindCSS
• Machine Learning: Stacking Classifier (Logistic Regression, Random Forest, LGBM)

---

📁 Project Structure

June Hackathon/
│
├── main.py               # FastAPI Backend Entry Point
├── requirements.txt      # Backend Dependencies
│
└── frontend/
    └── jobbot/          # Next.js Frontend App
        ├── package.json
        ├── tsconfig.json
        └── src/
            └── app/
                └── page/
                    ├── (multiple folders/files here for each page)

---

⚙️ Setup Instructions (Local Machine)

1. Clone the Repository

git clone [<repository-link>](https://github.com/Bhawesh-Agrawal/Numeric-Nomads)
cd June\ Hackathon

---

2. Backend Setup (FastAPI)

a) Create & Activate a Virtual Environment:

For Windows: Make sure python 3.12 is installed in your setup.

python3.12 -m venv venv
venv\Scripts\activate

For Linux/Mac:

python3.12 -m venv venv
source venv/bin/activate

b) Install Required Packages:

pip install -r requirements.txt

c) Run FastAPI Server:

uvicorn main:app --host 0.0.0.0 --port 8000

FastAPI Backend will now run at: 👉 http://localhost:8000

---

3. Frontend Setup (Next.js)

a) Go to Frontend Folder:

cd frontend/jobbot

b) Install Dependencies:

npm install

c) Change API URLs:

IMPORTANT: Go to each API calling file inside:

frontend/jobbot/src/app/page/...(files)

and replace the cloud URLs with your local FastAPI server URL:

http://localhost:8000

Do this wherever API requests are made (GET, POST, etc.).

d) Run Next.js Frontend:

npm run dev

Next.js Frontend will now run at: 👉 http://localhost:3000

---

4. Open the App:

• Backend: http://localhost:8000
• Frontend: http://localhost:3000

✅ You're all set to use the Job Fraud Detection System locally!

---

🚀 Our Process:

1. Problem Definition: Detect potential fraudulent job postings using machine learning.

2. Model Development: Trained a Stacking Classifier combining:

   • Logistic Regression
   • Random Forest
   • LightGBM with TF-IDF feature extraction.

3. Backend: FastAPI serves the trained model to make real-time predictions.

4. Frontend: Next.js-powered modern UI where users can:

   • Search for jobs
   • Check fraud probability
   • View detailed job descriptions

5. Integration: API endpoints from FastAPI consumed in Next.js app for seamless interaction.

6. Real Time Job: FastAPI also tracks real time job and calcualtes its fraud probability in real time.

---

👨‍💻 Done By:

Name	Role
Bhawesh Agrawal	Developer
Priyanka Singh	Team Lead

Team Name: Numeric Nomads Hackathon: Anveshan Hackathon 2025

---