⭐ Rate Platform - AI-Powered Review & Rating System

A comprehensive review and rating platform for businesses, services, and products with AI-powered fake review detection, sentiment analysis, and fraud prevention capabilities.

🏗️ System Architecture with AI Integration

graph TB
    %% User Interfaces
    UI["`**Next.js Frontend**<br/>localhost:3000<br/>- User Dashboard<br/>- Admin Interface<br/>- Report Management`"]
    EXT["`**Browser Extension**<br/>Chrome/Edge<br/>- Real-time Detection<br/>- Page Analysis<br/>- Quick Report`"]

    %% Core Platform
    STRAPI["`**Strapi CMS**<br/>localhost:1337<br/>- Content Management<br/>- API Gateway<br/>- Admin Panel<br/>- Authentication`"]

    %% AI Service
    AI["`**AI Analysis Service**<br/>GPT-5-mini<br/>- Scam Detection<br/>- Fake Review Detection<br/>- Sentiment Analysis<br/>- Batch Processing`"]

    %% Database
    DB[("`**PostgreSQL**<br/>Docker Container<br/>- User Data<br/>- Reports<br/>- AI Analysis Results<br/>- Validation Results`")]

    %% Services
    CRAWLER["`**Importer/Crawler**<br/>TypeScript Service<br/>- Web Scraping<br/>- Data Collection<br/>- FlareSolverr Proxy<br/>- AI Batch Processing`"]

    VALIDATOR["`**Validator Service**<br/>TypeScript/Python<br/>- Data Validation<br/>- Business Rules<br/>- Risk Scoring<br/>- Quality Check`"]

    %% External Services
    FLARE["`**FlareSolverr**<br/>Docker Service<br/>- Cloudflare Bypass<br/>- Proxy Service`"]

    REDIS["`**Redis**<br/>Cache & Queue<br/>- AI Result Cache<br/>- Job Processing<br/>- Inter-service Comm`"]

    %% User Flow
    USER(["`**Users**<br/>End Users<br/>Administrators`"])

    %% Connections
    USER --> UI
    USER --> EXT

    UI --> STRAPI
    EXT --> STRAPI

    STRAPI --> DB
    STRAPI --> VALIDATOR

    CRAWLER --> AI
    CRAWLER --> FLARE
    CRAWLER --> REDIS
    CRAWLER --> DB

    AI --> REDIS
    AI --> DB

    VALIDATOR --> REDIS
    VALIDATOR --> DB

    %% Styling
    classDef frontend fill:#e1f5fe
    classDef backend fill:#f3e5f5
    classDef database fill:#e8f5e8
    classDef service fill:#fff3e0
    classDef external fill:#fce4ec
    classDef ai fill:#fff9c4

    class UI,EXT frontend
    class STRAPI backend
    class DB database
    class CRAWLER,VALIDATOR service
    class FLARE,REDIS external
    class AI ai

Loading

🔄 Complete Data Processing Flow

1. Data Collection

Multiple Sources → Crawler/Importer → Raw Data Storage
     ↓                    ↓                 ↓
User Reviews         FlareSolverr      JSON Files
Business Data        Anti-captcha      Structured Data
Product Listings     Rate Limiting     Metadata

2. AI Processing Pipeline

Raw Data → AI Analysis (GPT-5-mini) → Enriched Data
    ↓              ↓                      ↓
Reviews      Fake Detection         Confidence Score
Ratings      Sentiment Analysis     Categories
Reports      Pattern Recognition    Severity Level

3. Validation & Storage

Enriched Data → Redis Queue → Validator → PostgreSQL
      ↓             ↓            ↓           ↓
AI Metadata    Async Jobs    Quality     Verified
Confidence     Pub/Sub       Scoring     Reviews
Categories     Caching       Dedup       Ratings

4. API Distribution

Strapi CMS → REST/GraphQL APIs → Client Applications
     ↓              ↓                    ↓
Content Mgmt    Endpoints          Next.js UI
Auth/Perms      Rate Limiting      Browser Ext
Admin Panel     Caching             Mobile Apps

📊 Key Features

Core Platform

• Review Management: Multi-category reviews (businesses, products, services)
• Rating System: 5-star ratings with weighted algorithms
• User Authentication: JWT-based auth with social login support
• Content Moderation: AI-powered content filtering and approval workflows

AI Capabilities (GPT-5-mini)

• Fake Review Detection: 80%+ accuracy in identifying fake reviews
• Sentiment Analysis: Understand customer emotions and feedback patterns
• Spam Filtering: Automatic detection of spam and promotional content
• Fraud Prevention: Identify scam patterns and suspicious activities
• Language Support: Vietnamese and English analysis

Technical Features

• Batch Processing: Handle 50+ reviews per batch (~12s processing)
• Redis Caching: 90%+ cache hit rate for duplicate detection
• Real-time Updates: WebSocket support for live notifications
• API Rate Limiting: Protect against abuse and ensure fair usage
• Data Validation: Multi-layer validation pipeline

🚀 Performance Metrics

• AI Processing Speed: 250 reviews/minute
• API Response Time: < 100ms (cached), < 500ms (uncached)
• Cache Hit Rate: 90%+ for duplicate content
• Uptime Target: 99.9% availability
• Cost Efficiency: ~$0.0003 per AI analysis (GPT-5-mini)

📚 Documentation

• Development Setup - Complete setup guide
• Modules Overview - Technical documentation for all modules
• API Documentation - Strapi API endpoints and schemas
• Frontend Guide - Next.js Web development guide
• AI Integration - AI service documentation

👀 Live demo

• UI - https://www.notum-dev.cz/
• Strapi - https://api.notum-dev.cz/admin
• Readonly user:
  • Email: [REDACTED]
  • Password: [REDACTED]

🥞 Tech Stack

Frontend

• Next.js v15 - React framework with App Router
• TailwindCSS v4 - Utility-first CSS framework
• Shadcn/ui - Beautiful UI components
• TypeScript - Type-safe development

Backend

• Strapi v5 - Headless CMS with REST & GraphQL
• PostgreSQL 17 - Primary database
• Redis - Caching and message queue
• Node.js 22 - Runtime environment

AI & Processing

• OpenAI GPT-5-mini - AI analysis engine
• FlareSolverr - Anti-bot bypass
• Playwright - Browser automation
• TypeScript/Python - Data processing

Infrastructure

• Turborepo - Monorepo management
• Docker - Containerization
• Yarn Workspaces - Package management

🚀 Getting started

Prerequisites

• Node.js 22.x - Required runtime
• Yarn 1.22.x - Package manager (enforced)
• PostgreSQL 17 - Database (or use Docker)
• Docker - For services (Redis, PostgreSQL, FlareSolverr)
• OpenAI API Key - For AI features (GPT-5-mini)

Quick Start

1. Clone and Install

git clone https://github.com/RateBox/Rate.git
cd Rate
yarn install

2. Environment Setup

# Copy environment templates
yarn setup:apps

# Configure AI service (root .env)
echo "OPENAI_API_KEY=your-api-key" >> .env
echo "OPENAI_MODEL=gpt-5-mini" >> .env
echo "AI_CACHE_ENABLED=true" >> .env

# Configure Strapi (apps/strapi/.env)
# Configure Next.js (apps/web/.env.local)

3. Start Services

# Start database (if not using local PostgreSQL)
cd apps/strapi && docker compose up -d db && cd ../..

# Start Redis for caching
docker run -d --name redis -p 6379:6379 redis:alpine

# Start all applications
yarn dev

4. Access Points

• Frontend: http://localhost:3000
• Admin Panel: http://localhost:1337/admin
• API: http://localhost:1337/api
• GraphQL: http://localhost:1337/graphql

AI Processing Workflow

# 1. Crawl data
cd apps/importer
ts-node-esm Scripts/crawl.ts links
ts-node-esm Scripts/crawl.ts crawl

# 2. Process with AI
ts-node-esm Scripts/ai-batch-processor.ts \
  -i ./Data/crawled \
  -o ./Data/processed

# 3. Push to validation
ts-node-esm Scripts/push-to-validation-with-ai.ts

📦 Project Structure

Apps

• apps/strapi - Strapi v5 CMS backend with custom plugins
• apps/web - Next.js v15 frontend application
• apps/importer - Data crawler and importer service

Packages

• packages/ai - AI analysis service (GPT-5-mini integration)
• packages/validator - TypeScript validation schemas
• packages/design-system - Shared TailwindCSS and CKEditor configs
• packages/eslint-config - Shared ESLint rules
• packages/prettier-config - Shared Prettier configuration
• packages/typescript-config - Shared TypeScript configurations

Modules

• Modules/Extension - Browser extension for real-time detection
• Modules/Validator - Python validation worker service

🛠️ Development Commands

# Development
yarn dev              # Start all services
yarn dev:strapi       # Start Strapi only
yarn dev:ui           # Start Next.js only

# Building
yarn build            # Build all apps
yarn build:strapi     # Build Strapi
yarn build:ui         # Build Next.js

# Code Quality
yarn lint             # Lint all code
yarn format           # Format with Prettier
yarn type-check       # TypeScript checking

# Testing
yarn test             # Run all tests
yarn test-plugin      # Test smart-component-filter plugin

🚨 Troubleshooting

Common Issues

Port already in use:

# Windows
netstat -ano | findstr :1337
taskkill /PID <PID> /F

# Linux/Mac
lsof -i :1337
kill -9 <PID>

Database connection failed:

# Check PostgreSQL is running
docker ps | grep postgres

# Restart database
docker restart rate-db

AI Service errors:

• Check OpenAI API key in .env
• Verify GPT-5-mini model access
• Check Redis is running for caching

📝 Contributing

1. Fork the repository
2. Create feature branch (git checkout -b feat/amazing-feature)
3. Commit changes (git commit -m 'feat: add amazing feature')
4. Push to branch (git push origin feat/amazing-feature)
5. Open Pull Request

Commit Convention

We use Conventional Commits:

• feat: New feature
• fix: Bug fix
• docs: Documentation
• style: Code style
• refactor: Code refactoring
• test: Tests
• chore: Maintenance

📄 License

MIT License - see LICENSE file for details

🙏 Acknowledgments

• Based on strapi-next-monorepo-starter
• AI powered by OpenAI GPT-5-mini
• UI components by Shadcn/ui