---
title: 🛡️ AdGuardian - AI Advertising Compliance Agent 🌍
emoji: 🛡️
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 4.44.0
app_file: app.py
pinned: true
license: mit
short_description: AI Agent for Advertising Compliance and Content Audit
models:
  - Qwen/Qwen3-8B
  - deepseek-ai/DeepSeek-V3-0324
  - deepseek-ai/DeepSeek-R1-0528
tags:
  - agent-demo-track
  - Agents-MCP-Hackathon
---

# 🛡️ AdGuardian - AI Advertising Compliance Agent

AdGuardian is an intelligent AI agent specialized in advertising compliance and marketing content auditing. It helps businesses ensure their advertising materials comply with regulatory standards across different countries while identifying potential legal issues before publication.

## 🎬 Demo Video

Watch AdGuardian in action - a quick demonstration of the three-stage analysis process:

[![AdGuardian Demo](https://img.shields.io/badge/Watch-Demo%20Video-FF0000?style=for-the-badge&logo=bilibili)](https://b23.tv/wwOrNLJ)

*Video showcases the three-stage analysis pipeline and key features of AdGuardian for advertising compliance.*

## 🎯 What Makes AdGuardian Special

AdGuardian employs a **three-stage progressive analysis** approach that mimics how professional advertising compliance auditors work:

1. **🔍 Stage 1: Basic Analysis** - Initial risk assessment and compliance evaluation
2. **🎨 Stage 2: Detailed Annotation** - Sentence-level risk marking with color-coded visualization
3. **✨ Stage 3: Revision Suggestions** - Corrected text with compliance improvements

## 🚀 Key Features

### 🧠 AI-Powered Intelligence
- **Real-time Streaming Analysis**: Watch AI think and analyze in real-time
- **Transparent Reasoning**: View complete AI thought processes (expandable sections)
- **Multiple AI Models**: Choose from free and limited-free premium models

### 🌍 Global Compliance Support
- **Multi-Country Standards**: USA, China, UK, Japan, Korea, Singapore, Malaysia
- **Multi-Language Analysis**: Support for English, Chinese, Japanese, Korean
- **Regulatory Expertise**: Built-in knowledge of advertising laws and standards

### 🎨 Professional Visualization
- **Color-Coded Risk Levels**: Green (safe) → Yellow (minor) → Orange (moderate) → Red (serious)
- **Interactive Interface**: Four-tab layout for easy navigation
- **HTML-Rendered Results**: Beautiful, professional output formatting

## 📋 How to Use

1. **Input Content**: Enter your marketing text or use provided examples
2. **Configure Settings**: Select target country, language, and AI model
3. **Start Analysis**: Click "Start AI Analysis" for real-time processing
4. **Review Results**: Navigate through four tabs:
   - **Stage 1**: Basic analysis with expandable AI reasoning
   - **Stage 2**: Color-coded sentence annotations
   - **Stage 3**: Before/after revision comparison
   - **Summary**: Complete overview and usage guide

## 🎪 Try the Examples

AdGuardian includes three carefully crafted examples showcasing common advertising issues:

- **🧴 Beauty Product (EN)**: Anti-aging serum with exaggerated claims
- **⚡ Energy Drink (EN)**: Performance beverage with unrealistic promises
- **🏠 Real Estate (中文)**: Property investment with guaranteed returns

Each example contains realistic but problematic content that demonstrates AdGuardian's analytical capabilities.

## 🤖 Supported AI Models

| Model | Provider | Type | Capabilities |
|-------|----------|------|-------------|
| Qwen/Qwen3-8B | SiliconFlow | Free | General analysis |
| DeepSeek-V3-0324 | Nebius | Limited Free | Advanced reasoning |
| DeepSeek-R1-0528 | Nebius | Limited Free | Enhanced analysis |
| Qwen3-235B-A22B | Nebius | Limited Free | Large-scale processing |
| Gemma-3-27b-it | Nebius | Limited Free | Vision + text analysis |

## 🛠️ Technical Architecture

- **Frontend**: Gradio with custom HTML rendering
- **AI Integration**: Multi-provider API support (SiliconFlow, Nebius)
- **Analysis Engine**: Three-stage progressive advertising compliance audit
- **Output Format**: Real-time streaming with professional visualization
- **Deployment**: Hugging Face Spaces with automatic scaling

## 🏆 Perfect for

- **Marketing Teams**: Ensure campaign compliance before launch
- **Legal Departments**: Pre-review advertising materials
- **Content Creators**: Avoid regulatory pitfalls
- **Agencies**: Streamline client content approval
- **Educators**: Teach advertising law and compliance

## 📄 License

MIT License

---

**AdGuardian**: Where AI meets advertising compliance, ensuring your content is both compelling and compliant. 🛡️✨