wrongNumber: RedFlag

AI-powered fraud detection for UPI scams, phishing messages, and financial fraud — built for the LPU Computing & AI Expo 2026.

---

What it does

Paste any suspicious message — a UPI alert, loan offer, KYC warning, or prize notification — and the system instantly classifies it as SAFE, SUSPICIOUS, or HIGH RISK (Fraud) using a machine learning model trained on Indian fraud datasets.

Detects:

• UPI & payment scams
• Phishing links
• Fake KYC / account suspension alerts
• Loan app fraud
• Prize / lottery scams
• OTP theft attempts

---

Architecture

Client (React)  →  Backend (Express)  →  ML Service (FastAPI)
    :5173              :3000                   :8000

Layer	Stack
Frontend	React 19, Vite, Tailwind CSS v4
Backend	Node.js, Express 5, Axios
ML Service	Python, FastAPI, scikit-learn, joblib

---

Project Structure

wrongNumber-redflag/
├── client/          # React frontend
├── backend/         # Express API gateway
└── ml-service/      # FastAPI + ML model
    ├── model.py          # Train the model
    ├── main.py           # FastAPI server
    ├── fraud_model.pkl   # Trained model (generated)
    └── vectorizer.pkl    # TF-IDF vectorizer (generated)

---

Getting Started

Prerequisites

• Node.js >= 18
• Python >= 3.10
• npm

---

1. ML Service

cd ml-service

# Create and activate virtual environment
python -m venv .venv
.venv\Scripts\activate          # Windows
# source .venv/bin/activate     # macOS/Linux

# Install dependencies
pip install fastapi uvicorn scikit-learn pandas joblib

# Train the model (generates fraud_model.pkl & vectorizer.pkl)
python model.py

# Start the ML service
uvicorn main:app --reload --port 8000

---

2. Backend

cd backend

# Install dependencies
npm install

# Create .env file
echo PORT=3000 > .env
echo ML_SERVICE=http://127.0.0.1:8000 >> .env

# Start the backend
npm run dev

---

3. Client

cd client

# Install dependencies
npm install

# Start the frontend
npm run dev

Open http://localhost:5173

---

API

POST /api/v1/analyze

Request

{ "message": "Your UPI account will be blocked. Verify KYC at bit.ly/upi-kyc" }

Response

{
  "result": {
    "prediction": "fraud",
    "confidence": 94.3,
    "risk_level": "DANGER"
  }
}

risk_level values: DANGER · WARNING · SAFE

---

Model

Trained with Logistic Regression + TF-IDF on a combined dataset of Indian SMS fraud and phishing messages.

• model.py — trains and saves the model
• main.py — serves predictions via FastAPI
• Re-train anytime: python model.py

---

Environment Variables

File	Variable	Description
backend/.env	PORT	Express server port (default: 3000)
backend/.env	ML_SERVICE	FastAPI base URL (default: http://127.0.0.1:8000)

---

wrongNumber: RedFlag — LPU Computing & AI Expo 2026