AdSageAI (RigtusHuddle)

AdSageAI is a cutting-edge, Multi-Agent Marketing System designed to optimize social media campaigns using advanced AI personas and real-time data. It leverages Google's Gemini models to simulate a diverse marketing team—Gen Z, Millennial, and Strategist—negotiating the best approach for your content.

---

🚀 Key Features

• Multi-Persona Agentic Core: Run parallel analysis using distinct AI personas (Youth vs. Adult) to get diverse feedback on a single campaign.
• "Human-in-the-Loop" Logic: A specialized Strategist Agent synthesizes conflicting viewpoints from other agents to propose a balanced, actionable strategy.
• Visual Predictive Analysis: Uses Gemini Vision (Multimodal) to "see" your draft images and predict engagement scores, emotional impact, and virality potential before you post.
• Dynamic Intelligence:
  • Real-time Hashtag Scraping: Automatically scrapes live, trending hashtags from the web based on AI-generated search queries (bypassing the knowledge cutoff of LLMs).
  • Structured Output: Enforces strict JSON schemas for UI rendering, turning unstructured AI thoughts into dynamic dashboards.

---

🛠️ Technical Architecture

Backend Stack

• Runtime: Python 3.12+
• Framework: Flask (REST API)
• AI Engine: Google Gemini 1.5 Flash / Pro (via google-genai SDK)
• Scraping: BeautifulSoup4 + requests (Targeting best-hashtags.com dynamically)
• Image Processing: Pillow (PIL)
• API Communication: RESTful endpoints with CORS enabled.

Frontend Stack

• Framework: Vanilla JavaScript (ES6+), HTML5, CSS3.
• Design: Custom CSS Grid/Flexbox layout (No heavy framework overhead).
• Communication: Fetch API for async non-blocking calls to the backend.

Project Structure

RigtusHuddle/
├── backend/
│   └── agent/
│       ├── server.py            # Flask API Entry Point
│       ├── agent_core.py        # Main Agent Orchestrator (The Brain)
│       ├── hashtag_scraper.py   # Real-time Web Scraper Tool
│       ├── prompts/             # System Instructions for Agents
│       │   ├── analyze_campaign.prompt (Youth Persona)
│       │   ├── analyze_campaign_adult.prompt (Adult Persona)
│       │   ├── negotiate_suggestions.prompt (Strategist)
│       │   └── predictive_analysis.prompt (Vision Agent)
│       └── tools/               # Helper modules (Image Gen, LinkedIn API)
├── UI/
│   ├── dashboard.html           # Main User Interface
│   ├── app.js                   # Client-side Logic
│   └── style.css                # Styling
└── requirements.txt             # Python Dependencies

---

📦 Installation & Setup

Prerequisites

• Python 3.10 or higher
• A Google Cloud API Key (for Gemini)

1. Clone the Repository

git clone https://github.com/alysajad/adsageai.git
cd adsageai/RigtusHuddle

2. Install Dependencies

It is recommended to use a virtual environment.

# Windows
python -m venv .venv
.venv\Scripts\activate

# Install requirements
pip install -r requirements.txt

3. Configure Environment

Create a .env file in the RigtusHuddle root directory:

GOOGLE_API_KEY=your_actual_api_key_here
LINKEDIN_CLIENT_ID=optional_for_posting
LINKEDIN_CLIENT_SECRET=optional_for_posting

---

🚦 Usage Guide

1. Start the Backend API

Navigate to the agent directory and run the Flask server:

cd backend/agent
python server.py
# Server runs on http://localhost:5000

2. Start the Frontend

Open a new terminal, navigate to the UI directory, and start a simple HTTP server:

cd ../../UI
python -m http.server 8000
# Access dashboard at http://localhost:8000/dashboard.html

3. Run a Campaign Analysis

1. Open http://localhost:8000/dashboard.html.
2. Text Campaign: Enter a description of your product or campaign idea.
   • System Action: The Youth and Adult agents will analyze it. The Strategist will merge their insights and display a strategy.
3. Draft Analysis: Upload an image and add a caption.
   • System Action: The Vision Agent will predict the "Engagement Score" (1-10) and suggest live hashtags. Note how the hashtags are fetched dynamically from the web!

---

🔌 API Reference

Endpoint	Method	Description
/analyze	POST	Triggers the multi-agent negotiation flow (Youth + Adult + Strategist).
/analyze_draft	POST	Uploads an image for predictive visual analysis & hashtag scraping.
/generate_image_for_post	POST	Generates a new image asset based on a text prompt.
/status	GET	Checks system health and API connections.

---

🤝 Contributing

1. Fork the repo.
2. Create your feature branch (git checkout -b feature/amazing-feature).
3. Commit your changes (git commit -m 'Add amazing feature').
4. Push to the branch (git push origin feature/amazing-feature).
5. Open a Pull Request.

---

Built with ❤️ for the AI Agent Hackathon.