Satarkta - AI-Powered Invoice Fraud Detection Platform

Overview

Satarkta is a comprehensive fraud detection platform that uses AI and machine learning to analyze invoices and detect potential fraud indicators. The platform now includes bank account verification capabilities using penny-less verification.

Features

Core Features

• AI-Powered Invoice Analysis: Advanced OCR and ML algorithms for invoice processing
• GSTIN Verification: Real-time verification against government databases
• Duplicate Detection: Machine learning-based duplicate invoice detection
• Compliance Checking: GSTR-2B reconciliation and e-invoice IRN validation
• Risk Scoring: Intelligent fraud risk assessment (0-100 scale)

New Features

• Bank Account Verification: Penny-less bank account verification using Sandbox API
• IFSC Code Validation: Real-time IFSC code verification with bank details
• Enhanced Dashboard: Tabbed interface with multiple verification tools
• Optimized API: RESTful API endpoints optimized for frontend integration

Architecture

Backend (Python Flask)

• optimized_app.py: Main Flask application with enhanced API endpoints
• bank_verification.py: Bank account verification module
• invoice_utils.py: Invoice processing and OCR utilities
• compliance_utils.py: GST compliance checking
• gstin_utils.py: GSTIN verification utilities
• duplicate_detection.py: ML-based duplicate detection

Frontend (Next.js)

• Enhanced Dashboard: Multi-tab interface for different verification tools
• Bank Verification Component: User-friendly bank account verification
• Real-time Processing: Optimized API integration
• Responsive Design: Mobile-friendly interface

Setup Instructions

Prerequisites

• Python 3.8+
• Node.js 16+
• AWS Account (for S3 storage)

Backend Setup

1. Navigate to backend directory:

   cd backend

2. Install dependencies:

   pip install -r requirements.txt

3. Create .env file with required variables:

   TELEGRAM_TOKEN=your_telegram_bot_token
   S3_BUCKET=your_s3_bucket_name
   SANDBOX_API_KEY=your_sandbox_api_key
   SANDBOX_API_SECRET=your_sandbox_api_secret
   AWS_ACCESS_KEY_ID=your_aws_access_key
   AWS_SECRET_ACCESS_KEY=your_aws_secret_key

4. Run the backend:

   python optimized_app.py

Frontend Setup

1. Navigate to frontend directory:

   cd frontend

2. Install dependencies:

   npm install

3. Run the development server:

   npm run dev

Running Both Services

You can run both backend and frontend simultaneously:

cd frontend
npm run dev:all

API Endpoints

Invoice Processing

• POST /api/process-invoice - Process uploaded invoice PDF
• GET /api/recent-scans - Get recent invoice scans
• GET /api/dashboard-stats - Get dashboard statistics

Bank Verification

• POST /api/verify-bank-account - Verify bank account details
• POST /api/verify-ifsc - Verify IFSC code

System

• GET /api/health - Health check endpoint

Usage

Invoice Processing

1. Upload PDF invoice through the dashboard
2. System extracts key information using OCR
3. Performs multiple fraud checks:
   • GSTIN verification
   • Duplicate detection
   • Compliance checking
   • Risk scoring
4. Returns detailed analysis with recommendations

Bank Account Verification

1. Enter account number and IFSC code
2. Optionally provide account holder name for matching
3. System verifies account existence and details
4. Returns verification status and bank information

Deployment

Without LocalTunnel

The optimized backend runs on localhost:5001 and the frontend on localhost:3000. For production deployment:

1. Backend: Deploy to cloud service (AWS EC2, Google Cloud, etc.)
2. Frontend: Build and deploy to static hosting (Vercel, Netlify, etc.)
3. Update API URLs: Change frontend API calls to production backend URL

Environment Variables for Production

# Backend
FLASK_ENV=production
PORT=5001

# Frontend
NEXT_PUBLIC_API_URL=https://your-backend-domain.com

Security Features

• Input validation and sanitization
• Rate limiting on API endpoints
• Secure file upload handling
• Data encryption in transit
• Audit trail for all operations

Monitoring and Analytics

• Real-time fraud detection metrics
• Historical analysis and trends
• Risk distribution analytics
• Performance monitoring

Support

For technical support or questions, contact: support@satarkta.in

License

Proprietary - All rights reserved