PayGuard AI

AI-powered transaction risk monitoring, compliance automation, and fraud detection for cross-border payments

🌐 Live Demo: https://payguard-ai-production.up.railway.app

Overview

PayGuard AI is a compliance and risk tooling SaaS platform built for the Afriex Cross-Border Fintech Hackathon. It provides financial institutions with real-time transaction monitoring, ML-powered risk scoring, a Human-in-the-Loop (HITL) review workflow, multi-tenancy, and a Rule Marketplace — all deployed on Railway with PostgreSQL.

The Problem

Cross-border payment platforms process thousands of transactions daily. Compliance teams struggle with:

Manual review bottlenecks — Can't keep up with transaction volume
Delayed fraud detection — Suspicious activity caught too late
Inconsistent decisions — No standardized risk assessment framework
Audit trail gaps — Difficulty proving compliance to regulators
One-size-fits-all rules — No industry-specific tuning for risk thresholds

The Solution

PayGuard AI automates the first line of defense while keeping humans in control of critical decisions:

Real-time Risk Scoring — Every transaction analyzed instantly via configurable rules + ML model
ML-Powered Fraud Detection — Learns from HITL feedback, auto-retrains hourly
Smart Routing — Low-risk transactions auto-approve; high-risk ones go to human reviewers
Human-in-the-Loop — Compliance officers review, approve, or reject flagged transactions
Rule Marketplace — Pre-built industry packs (Remittance, E-Commerce, Lending, Crypto) with one-click import
Complete Audit Trail — Every decision logged for regulatory compliance
Multi-Tenancy — Full data isolation per organization with RBAC

Features

Core Platform

📊 Live Dashboard — Real-time stats, charts, risk distribution, and transaction monitoring
🔍 Risk Analysis Engine — 6 configurable rules + ML scoring with per-rule analytics
👥 HITL Review Queue — Prioritized list of transactions needing human review
⚡ Real-time Updates — SignalR-powered instant notifications
📋 Rules Management — Create, edit, toggle, and import risk detection rules
📈 Compliance Reports — Visual analytics with risk distribution charts and CSV export
📝 Audit Logging — Complete history of all actions and decisions

AI & Machine Learning

🤖 ML Risk Scoring — Binary classification model trained on HITL feedback (FastTree)
🔄 Auto-Retraining — Background service checks hourly for new labeled data
📊 Model Management — View training metrics (AUC, F1, precision/recall), activate/deactivate models
🧠 Feature Engineering — 12 features extracted from transaction context (amount, velocity, time, corridor risk)
💡 Smart Rule Suggestions — ML-driven analysis of review patterns to suggest new rules and threshold adjustments

Rule Marketplace

🏪 Template Catalog — 24 pre-built templates across 4 industries
📦 Industry Packs — One-click import of all 6 rules optimized for your industry
📊 Rule Analytics — Per-rule effectiveness: hit rate, precision, false positive rate
🔄 Import/Update — Import new rules or update existing ones with recommended thresholds
📜 Rule Versioning — Full version history with diff comparison and rollback

Fraud Detection & Intelligence

🕸️ Fan-out/Fan-in Detection — Network analysis to detect structuring rings where one sender splits to many receivers or many senders funnel to one receiver
🚫 Watchlists & Blocklists — Custom watchlists with automatic matching against transactions using name/email/country criteria
🔗 Relationship Analysis — Graph-based visualization of transaction networks between entities

Compliance & Reporting

🛡️ GDPR Compliance — Data subject search, export (JSON/CSV), right-to-erasure with full audit trail
📄 Invoice PDF Generation — QuestPDF-powered professional invoices with automatic numbering and PDF download
📊 Advanced Reports — Scheduled report generation with background processing and viewer dialog
🧾 System Logs — Centralized, structured logging with retention policies and level-based filtering

Enterprise Features

🏢 Multi-Tenancy — Tenant-scoped data isolation via middleware + EF Core query filters
🔐 OAuth 2.0 & Magic Links — Production-ready auth (Azure AD/Google/Okta) + passwordless login
👮 RBAC — 4-tier roles: Reviewer, Manager, Admin, SuperAdmin with custom permissions
🚀 Tenant Onboarding — Guided wizard for new organizations
💳 Billing — Paystack-powered subscription management with usage-based pricing tiers (Trial/Starter/Pro/Enterprise)
📧 Email Notifications — Resend-powered alerts for critical risk events with per-user preferences
🔑 API Keys & Webhooks — Self-service API key management and webhook configuration with signature verification
💱 Multi-Provider Support — Afriex, Flutterwave, Wise payment provider abstraction

Operations & Monitoring

📡 Monitoring Dashboard — Real-time operational metrics: throughput, error rates, risk distribution, review queue depth, webhook activity, and 7-day trends
🚦 Rate Limiting — Fixed-window rate limiter scoped per tenant
💾 Response Caching — In-memory caching for dashboard stats and transactions
🚨 Alerting Service — Automatic alerts for critical-risk transactions
📡 Health Checks — /health endpoint for uptime monitoring
📊 Prometheus Metrics — /metrics endpoint with request timing and slow-request warnings
🚩 Feature Flags — Safe deployment with instant rollback
🐘 PostgreSQL — Production database on Railway (SQLite for local dev)
🔒 Security Hardened — No eval() injection, Swagger restricted to dev, secure cookie policies, sanitized error messages

Mobile & PWA

📱 Progressive Web App — Installable on mobile with offline shell caching
🔽 Mobile Bottom Navigation — Touch-friendly nav bar with badge counts
👆 Swipe-to-Review — Swipe right to approve, left to reject on mobile review queue
🔄 Pull-to-Refresh — Touch-native refresh gesture on review and transaction lists
📐 Responsive Layout — Auto-closing drawer on mobile, compact cards, 48px touch targets
🔔 Web Push Notifications — Service worker push support for real-time alerts

Tech Stack

Layer	Technology
Frontend	Blazor Server, MudBlazor 8.x
Backend	ASP.NET Core 10
Database	PostgreSQL (production) / SQLite (development)
ML	ML.NET (FastTree binary classification)
Real-time	SignalR WebSockets
Auth	OAuth 2.0 / Magic Links / Demo mode
Email	Resend HTTP API
Billing	Paystack
PDF	QuestPDF
Providers	Afriex, Flutterwave, Wise
Caching	IMemoryCache (tenant-scoped)
Monitoring	Prometheus, Health Checks, Serilog
Deployment	Railway (Docker)
Architecture	Clean Architecture (3-layer)

Project Structure

PayGuardAI/
├── src/
│   ├── PayGuardAI.Core/                  # Domain entities and interfaces
│   │   ├── Entities/                      # 23 entities
│   │   │   ├── Transaction.cs             # Transaction entity
│   │   │   ├── RiskAnalysis.cs            # Risk scoring results
│   │   │   ├── RiskRule.cs                # Configurable risk rules
│   │   │   ├── RuleTemplate.cs            # Marketplace templates
│   │   │   ├── RuleVersion.cs             # Rule version history
│   │   │   ├── RuleGroup.cs               # Compound rule groups
│   │   │   ├── MLModel.cs                 # ML model storage
│   │   │   ├── CustomerProfile.cs         # Customer risk profiles
│   │   │   ├── AuditLog.cs                # Audit trail
│   │   │   ├── SystemLog.cs               # Centralized system logs
│   │   │   ├── TeamMember.cs              # RBAC team members
│   │   │   ├── CustomRole.cs              # Custom permission roles
│   │   │   ├── Invoice.cs                 # Billing invoices
│   │   │   ├── Watchlist.cs               # Watchlists & blocklists
│   │   │   ├── WatchlistEntry.cs          # Watchlist entries
│   │   │   ├── WebhookEndpoint.cs         # Webhook configuration
│   │   │   ├── TenantSubscription.cs      # Billing subscriptions
│   │   │   ├── OrganizationSettings.cs    # Tenant settings
│   │   │   └── ...                        # ApiKey, MagicLinkToken, etc.
│   │   └── Services/                      # 23 service interfaces
│   │       ├── IRiskScoringService.cs
│   │       ├── IRuleMarketplaceService.cs
│   │       ├── IRuleSuggestionService.cs
│   │       ├── IMLScoringService.cs
│   │       ├── IWatchlistService.cs
│   │       ├── IRelationshipAnalysisService.cs
│   │       ├── IGdprService.cs
│   │       ├── ITenantContext.cs
│   │       └── ...
│   │
│   ├── PayGuardAI.Data/                  # Data access and service implementations
│   │   ├── ApplicationDbContext.cs        # EF Core context with multi-tenant query filters
│   │   └── Services/                      # 34 service implementations
│   │       ├── RiskScoringService.cs          # Rule evaluation + ML scoring
│   │       ├── RuleMarketplaceService.cs      # Template browsing, import, analytics
│   │       ├── RuleSuggestionService.cs       # ML-driven rule suggestions
│   │       ├── RuleVersioningService.cs       # Rule version tracking
│   │       ├── MLScoringService.cs            # ML prediction engine
│   │       ├── MLTrainingService.cs           # Model training pipeline
│   │       ├── TransactionService.cs          # Cached, tenant-scoped
│   │       ├── ReviewService.cs               # HITL review workflow
│   │       ├── WatchlistService.cs            # Watchlist matching
│   │       ├── RelationshipAnalysisService.cs # Fan-out/fan-in detection
│   │       ├── GdprService.cs                 # GDPR data operations
│   │       ├── InvoiceService.cs              # Invoice CRUD
│   │       ├── MonitoringService.cs           # Real-time operational metrics
│   │       ├── TenantOnboardingService.cs     # Guided tenant setup
│   │       ├── DatabaseMigrationService.cs    # Auto-migration for PostgreSQL/SQLite
│   │       ├── WebhookDeliveryService.cs      # Webhook dispatch + retry
│   │       └── ...
│   │
│   └── PayGuardAI.Web/                   # Blazor UI, API controllers, middleware
│       ├── Components/Pages/              # 47 pages/dialogs
│       │   ├── Home.razor                 # Dashboard with live stats
│       │   ├── Transactions.razor         # Transaction list with filters
│       │   ├── Reviews.razor              # HITL review queue
│       │   ├── Rules.razor                # Rule management + suggestions
│       │   ├── RuleMarketplace.razor      # Template browsing + analytics
│       │   ├── MLModels.razor             # ML model management
│       │   ├── Reports.razor              # Compliance analytics + CSV export
│       │   ├── Audit.razor                # Audit log viewer
│       │   ├── Send.razor                 # Transaction simulator
│       │   ├── NetworkAnalysis.razor      # Fan-out/fan-in graph visualization
│       │   ├── Watchlists.razor           # Watchlist management
│       │   ├── GdprCompliance.razor       # GDPR search, export, erasure
│       │   ├── Invoices.razor             # Invoice management + PDF download
│       │   ├── Monitoring.razor           # Operational monitoring dashboard
│       │   ├── SystemLogs.razor           # Centralized log viewer
│       │   └── ...                        # Billing, Profile, Settings, etc.
│       ├── Controllers/                   # 6 API controllers
│       │   ├── WebhooksController.cs      # Multi-provider webhooks
│       │   ├── AuthController.cs          # Auth endpoints (OAuth, magic link, demo)
│       │   ├── InvoiceController.cs       # PDF download endpoint
│       │   └── ...
│       ├── Services/                      # 20 middleware & background services
│       │   ├── TenantResolutionMiddleware.cs
│       │   ├── SecurityHeadersMiddleware.cs
│       │   ├── InputValidationMiddleware.cs
│       │   ├── MLRetrainingBackgroundService.cs
│       │   ├── ScheduledReportBackgroundService.cs
│       │   ├── LogRetentionBackgroundService.cs
│       │   ├── InvoicePdfService.cs
│       │   └── ...
│       ├── Hubs/
│       │   └── TransactionHub.cs          # SignalR real-time hub
│       └── wwwroot/
│           └── js/payguard.js             # Safe JS interop helpers
│
└── tests/
    └── PayGuardAI.Tests/                  # 266 tests
        ├── Services/                       # 10 unit test classes
        │   ├── RuleMarketplaceServiceTests.cs
        │   ├── TenantOnboardingTests.cs
        │   ├── RbacServiceTests.cs
        │   ├── MLFeatureExtractorTests.cs
        │   ├── SecurityMiddlewareTests.cs
        │   ├── AfriexProviderTests.cs
        │   ├── FlutterwaveProviderTests.cs
        │   ├── WiseProviderTests.cs
        │   ├── PaymentProviderFactoryTests.cs
        │   └── TenantIsolationTests.cs
        └── Integration/                    # API integration tests
            ├── WebhooksControllerIntegrationTests.cs
            └── SecurityIntegrationTests.cs

Getting Started

Prerequisites

.NET 10 SDK
Git
Docker Desktop (optional, for containerized deployment)

Installation

Option 1: Local Development (Recommended)

# Clone the repository
git clone https://github.com/Karinateii/PayGuard-AI.git
cd PayGuard-AI

# Restore dependencies
dotnet restore

# Run the application (uses SQLite by default)
cd src/PayGuardAI.Web
dotnet run

# Open in browser
open http://localhost:5054

Option 2: Docker (Production)

# Clone and start with Docker Compose
git clone https://github.com/Karinateii/PayGuard-AI.git
cd PayGuard-AI
./start-docker.sh
# Or: docker-compose up -d

# Open in browser
open http://localhost:5054

# View logs / stop
docker-compose logs -f payguard-web
docker-compose down

The app comes with 25 demo transactions and 24 rule templates pre-seeded for testing.

Deployment

For production deployment to Railway, Heroku, or other cloud platforms, see DEPLOYMENT.md and DOCKER-HEROKU-GUIDE.md.

Authentication & Security

PayGuard AI supports three authentication modes:

Development Mode (Default)

Demo authentication is enabled by default for quick testing:

Access the dashboard immediately at http://localhost:5054
Demo user: compliance_officer@payguard.ai (SuperAdmin)
No login required

Magic Link (Passwordless)

Passwordless email authentication via magic links:

Users receive a one-time login link via email (Resend API)
No passwords to manage or forget

Production Mode (OAuth 2.0)

Enable enterprise-grade authentication by setting FeatureFlags:OAuthEnabled to true:

{
  "FeatureFlags": { "OAuthEnabled": true },
  "OAuth": {
    "Provider": "AzureAD",
    "TenantId": "your-tenant-id",
    "ClientId": "your-client-id",
    "ClientSecret": "your-client-secret"
  }
}

Supported Providers: Azure AD, Google Workspace, Okta, any OIDC provider.

Security Hardening

No eval() injection — All JS interop uses safe, parameterised helper functions (wwwroot/js/payguard.js)
Swagger restricted to development — API docs are not exposed in production
Secure cookies — HttpOnly, SameSite=Lax, SecurePolicy=Always
Sanitized error messages — Exception details never leak to the UI; generic errors shown to users with full stack traces logged server-side
Security headers — CSP, X-Content-Type-Options, X-Frame-Options, Referrer-Policy via middleware
Input validation — Request validation middleware rejects malformed payloads
Webhook signature verification — HMAC-based verification for all inbound webhooks

Role-Based Access Control

Role	Access Level
Reviewer	View transactions, approve/reject flagged items, view reports
Manager	+ Rules, Billing, Invoices, Audit, Rule Marketplace, Watchlists
Admin	+ Team, API Keys, Webhooks, Analytics, ML Models, Organization Settings, Monitoring, System Logs, GDPR
SuperAdmin	+ Tenant Management (platform owner)

Multi-Provider Integration

PayGuard AI supports multiple payment providers through a unified abstraction layer:

Provider	Endpoint	Status
Afriex	`POST /api/webhooks/afriex`	✅ Always enabled
Flutterwave	`POST /api/webhooks/flutterwave`	⚙️ Feature flag
Wise	`POST /api/webhooks/wise`	⚙️ Feature flag

Testing Webhooks

# Afriex
curl -X POST http://localhost:5054/api/webhooks/afriex \
  -H "Content-Type: application/json" \
  -H "X-Afriex-Signature: test-signature" \
  -d '{"event":"transaction.completed","data":{"id":"TEST-001","type":"send","status":"completed","amount":500,"currency":"USD","source_country":"US","destination_country":"NG","customer":{"id":"cust-001","email":"test@example.com","name":"Test User"},"created_at":"2026-02-26T10:00:00Z"}}'

Risk Scoring

Rule-Based Scoring (6 configurable rules)

Rule Code	Description	Default Threshold
`HIGH_AMOUNT`	Large transaction amount	$5,000
`VELOCITY_24H`	Too many transactions in 24h	5 txns
`NEW_CUSTOMER`	First-time or new customer	< 3 txns
`HIGH_RISK_CORRIDOR`	OFAC-sanctioned countries	IR, KP, SY, YE, VE, CU
`ROUND_AMOUNT`	Suspiciously round amounts	$1,000
`UNUSUAL_TIME`	Transactions at 2–5 AM UTC	Always flags

ML Risk Scoring

The ML model augments rule-based scoring with learned patterns:

Algorithm: FastTree binary classification (ML.NET)
Features: 12 dimensions including amount, velocity, time, corridor risk, customer history
Training: Learns from HITL review decisions (Approved = legitimate, Rejected = fraud)
Auto-retraining: Background service checks hourly, retrains when 50+ new labeled samples exist
Model management: View metrics, compare versions, activate/deactivate from Admin panel

Smart Rule Suggestions

The platform analyzes review patterns and transaction data to automatically suggest:

New rules based on frequently-rejected transaction characteristics
Threshold adjustments when existing rules under- or over-flag
One-click apply to immediately enable suggested rules

Risk Levels

Level	Score Range	Action
Low	0–25	Auto-approved
Medium	26–50	Flagged for review
High	51–75	Requires manual review
Critical	76–100	Requires manual review

Rule Marketplace

Pre-built rule templates optimized for different industries:

Industry	Templates	Example Threshold
Remittance	6 rules	HIGH_AMOUNT: $10,000, VELOCITY: 3/day
E-Commerce	6 rules	HIGH_AMOUNT: $2,000, VELOCITY: 15/day
Lending	6 rules	HIGH_AMOUNT: $5,000, VELOCITY: 2/day
Crypto	6 rules	HIGH_AMOUNT: $50,000, VELOCITY: 10/day

Features:

Browse and filter by industry, category, or keyword
One-click import of individual templates or entire industry packs
Rule analytics with precision, hit rate, and false positive tracking
Import count (popularity) tracking across tenants

Fraud Intelligence

Fan-out / Fan-in Detection

Network graph analysis identifies structuring rings:

Fan-out: One sender splitting transactions across many receivers to stay below thresholds
Fan-in: Many senders funnelling money to a single receiver
Interactive graph visualization on the Network Analysis page
Configurable thresholds and time windows

Watchlists & Blocklists

Create custom watchlists with name, email, and country criteria
Automatic real-time matching against incoming transactions
Manual override options for compliance officers
Bulk import/export support

GDPR Compliance

Full General Data Protection Regulation tooling:

Data Subject Search — Find all data for a customer by email or name
Data Export — One-click export of all customer data in JSON or CSV format
Right to Erasure — Anonymize or delete customer data with confirmation dialog
Audit Trail — Every GDPR action is logged for regulatory proof

Monitoring & Observability

Operational Dashboard (`/admin/monitoring`)

Real-time operational health with 30-second auto-refresh:

Health Banner — Healthy / Warning / Degraded status based on error rate
Throughput Metrics — 24h transaction count with hourly breakdown chart
Risk Distribution — Donut chart of risk level distribution
7-Day Trend — Daily transaction volume bar chart
Error Rate — Percentage of error-level system logs
Review Queue — Pending review count for capacity planning
Webhook Activity — Delivery success/failure rates
Active Rules — Count of enabled risk detection rules

System Logs (`/admin/logs`)

Centralized structured logging via Serilog
Filter by level (Debug, Info, Warning, Error, Fatal), source, and date range
Automatic log retention with configurable cleanup via background service

Endpoints

Endpoint	Auth	Description
`/health`	Public	Application health check
`/metrics`	Admin+	Prometheus metrics

Invoice & Billing

Subscription Tiers

Plan	Price	Transactions/mo
Trial	$0	100
Starter	$99/mo	1,000
Pro	$499/mo	10,000
Enterprise	$2,000/mo	Unlimited

Invoice PDF Generation

Automatic invoice numbering (INV-YYYY-NNNN)
Professional A4 PDF layout generated with QuestPDF
Download via API endpoint (GET /api/invoices/{id}/pdf)
Invoice history with summary cards (total billed, outstanding, overdue)

Testing

# Run all 266 tests
dotnet test

# Run with detailed output
dotnet test --logger "console;verbosity=detailed"

# Run specific test class
dotnet test --filter "RuleMarketplaceServiceTests"

Test Coverage

Test Class	Tests	Coverage
PaymentProviderFactoryTests	48	Factory pattern, provider selection
AfriexProviderTests	30	Afriex API integration
FlutterwaveProviderTests	28	Flutterwave normalization
RuleMarketplaceServiceTests	25	Template browsing, import, analytics
RbacServiceTests	24	Roles, permissions, team management
MLFeatureExtractorTests	20	Feature extraction for ML
WiseProviderTests	20	Wise transfer mapping
TenantOnboardingTests	16	Tenant setup, rule seeding
SecurityMiddlewareTests	15	Auth, rate limiting, CORS
TenantIsolationTests	—	Multi-tenant data isolation
Integration Tests	40	End-to-end webhook processing
Total	266

Continuous Integration

GitHub Actions workflow runs on every push:

✅ Multi-platform testing (Ubuntu, Windows, macOS)
✅ Code quality checks
✅ Security vulnerability scanning

API Endpoints

Method	Endpoint	Auth	Description
POST	`/api/webhooks/afriex`	API Key	Receive Afriex transaction webhooks
POST	`/api/webhooks/flutterwave`	API Key	Receive Flutterwave webhooks
POST	`/api/webhooks/wise`	API Key	Receive Wise webhooks
POST	`/api/auth/demo-login`	Anonymous	Demo login (dev mode only)
POST	`/api/auth/magic-link`	Anonymous	Request magic link email
GET	`/api/auth/verify`	Anonymous	Verify magic link token
GET	`/api/invoices/{id}/pdf`	Manager+	Download invoice PDF
GET	`/health`	Public	Application health check
GET	`/metrics`	Admin+	Prometheus metrics
—	`/transactionHub`	Authenticated	SignalR real-time connection

Multi-Tenancy

Each organization gets fully isolated data:

Middleware-based resolution: X-Tenant-Id header or email→tenant lookup
EF Core query filters: All queries automatically scoped to current tenant
Tenant onboarding: Guided wizard seeds rules, settings, and team
Default tenant: afriex-demo for development

{
  "MultiTenancy": {
    "DefaultTenantId": "afriex-demo"
  }
}

Database

Local Development (SQLite)

# Reset database
rm src/PayGuardAI.Web/payguardai.db
dotnet run  # Auto-recreates with seed data

Production (PostgreSQL)

PostgreSQL is enabled via feature flag. The DatabaseMigrationService automatically:

Creates all tables if they don't exist
Adds missing columns to existing tables
Seeds default data (rules, templates, demo transactions)
Fixes indexes for multi-tenancy

{
  "FeatureFlags": { "PostgresEnabled": true },
  "ConnectionStrings": {
    "PostgreSQL": "Host=...;Database=payguard;Username=...;Password=..."
  }
}

Configuration

Key settings in appsettings.json:

{
  "FeatureFlags": {
    "OAuthEnabled": false,
    "PostgresEnabled": false,
    "FlutterwaveEnabled": false,
    "WiseEnabled": false
  },
  "Auth": {
    "DefaultUser": "compliance_officer@payguard.ai",
    "DefaultRoles": "Reviewer,Manager,Admin,SuperAdmin"
  },
  "RateLimiting": {
    "PermitLimit": 60,
    "WindowSeconds": 60
  },
  "Afriex": {
    "BaseUrl": "https://staging.afx-server.com",
    "ApiKey": "your-api-key"
  }
}

Hackathon Track

Compliance and Risk Tooling — Building tools that help fintech companies maintain regulatory compliance while processing cross-border payments efficiently.

Contributing

This project was built for the Afriex Cross-Border Fintech Hackathon. Contributions, issues, and feature requests are welcome!

License

MIT License — see LICENSE file for details.

Acknowledgments

Afriex for the hackathon opportunity and API documentation
MudBlazor for the Blazor component library
ML.NET for the machine learning framework
QuestPDF for the PDF generation library
The ASP.NET Core team for SignalR and the middleware pipeline

Built with ❤️ for the Cross-Border Fintech Hackathon 2026

Name		Name	Last commit message	Last commit date
Latest commit History 215 Commits
.github		.github
scripts		scripts
src		src
tests/PayGuardAI.Tests		tests/PayGuardAI.Tests
.dockerignore		.dockerignore
.gitignore		.gitignore
DOCKER-HEROKU-GUIDE.md		DOCKER-HEROKU-GUIDE.md
Dockerfile		Dockerfile
PHASE-1-VERIFICATION-CHECKLIST.md		PHASE-1-VERIFICATION-CHECKLIST.md
POSTGRESQL-MIGRATION.md		POSTGRESQL-MIGRATION.md
PayGuardAI.slnx		PayGuardAI.slnx
README.md		README.md
docker-compose.yml		docker-compose.yml
heroku.yml		heroku.yml
start-docker.sh		start-docker.sh

Karinateii/PayGuard-AI

Folders and files

Latest commit

History

Repository files navigation

PayGuard AI

Overview

The Problem

The Solution

Features

Core Platform

AI & Machine Learning

Rule Marketplace

Fraud Detection & Intelligence

Compliance & Reporting

Enterprise Features

Operations & Monitoring

Mobile & PWA

Tech Stack

Project Structure

Getting Started

Prerequisites

Installation

Option 1: Local Development (Recommended)

Option 2: Docker (Production)

Deployment

Authentication & Security

Development Mode (Default)

Magic Link (Passwordless)

Production Mode (OAuth 2.0)

Security Hardening

Role-Based Access Control

Multi-Provider Integration

Testing Webhooks

Risk Scoring

Rule-Based Scoring (6 configurable rules)

ML Risk Scoring

Smart Rule Suggestions

Risk Levels

Rule Marketplace

Fraud Intelligence

Fan-out / Fan-in Detection

Watchlists & Blocklists

GDPR Compliance

Monitoring & Observability

Operational Dashboard (/admin/monitoring)

System Logs (/admin/logs)

Endpoints

Invoice & Billing

Subscription Tiers

Invoice PDF Generation

Testing

Test Coverage

Continuous Integration

API Endpoints

Multi-Tenancy

Database

Local Development (SQLite)

Production (PostgreSQL)

Configuration

Hackathon Track

Contributing

License

Acknowledgments

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Operational Dashboard (`/admin/monitoring`)

System Logs (`/admin/logs`)

Packages