Transform your ideas into platform-perfect ads with AI-powered multimodal generation. Create stunning text, images, and videos optimized for every platform—all in seconds, completely free.
- Multi-Platform Support: Generate ads for Instagram, LinkedIn, Twitter, and YouTube
- Agentic RAG System: Advanced Retrieval-Augmented Generation with LangGraph orchestration
- Multi-Agent Architecture: Specialized agents for research, copywriting, visual design, and quality assurance
- Knowledge Base Integration: Vector database with ChromaDB for enhanced content generation
- Advanced Admin Dashboard: Monitor usage, feedback, and analytics with multi-page navigation
- Real-time Analytics: Live data visualization with interactive charts and business intelligence
- Responsive Design: Works seamlessly on desktop and mobile devices
- Secure Admin Access: JWT-protected admin APIs with session storage
- Feedback-Driven Improvement Loop: Continuous learning and adaptation based on user feedback
- Framework: React 18 with TypeScript
- Build Tool: Vite
- UI Components: Custom components with CSS Modules
- Icons: Lucide React
- State Management: React Hooks with Context API
- Styling: CSS3 with CSS Variables and Glass Morphism effects
- Charts: Custom SVG-based visualizations
- Framework: FastAPI (Python 3.10+)
- Database: MongoDB with Motor (async driver)
- Authentication: JWT with bcrypt password hashing
- API: RESTful architecture with WebSocket support
- Orchestration: LangGraph for multi-agent workflows
- LLM Framework: LangChain
- Models: Hugging Face Transformers
- Vector Database: ChromaDB with FAISS for similarity search
- Knowledge Base: RAG (Retrieval-Augmented Generation) system
- Content Researcher: Gathers and analyzes information
- Copywriter Agent: Generates compelling ad copy
- Visual Designer: Creates engaging visual content
- Video Scriptwriter: Produces video scripts and storyboards
- Quality Assurance: Ensures brand consistency and quality
- Backend: Python 3.10 or higher (required for AI/ML dependencies)
- Frontend: Node.js 16+, npm or yarn
- Database: MongoDB (local or Atlas cluster)
- Vector Database: ChromaDB for RAG system
- Hugging Face API Key: For accessing AI models
- Create a
.envfile in the root directory with the following variables:
# MongoDB Configuration
MONGODB_URI=mongodb://localhost:27017/agenticads
# JWT Configuration
JWT_SECRET=your_jwt_secret_here
JWT_EXPIRE=24h
# Hugging Face
HUGGINGFACE_API_KEY=your_huggingface_api_key
# Vector Database
CHROMA_DB_PATH=./chroma_db
# Frontend
npm install
npm run dev
# Backend (Python 3.10+ required)
cd backend
pip install -r requirements.txt
python main.pyNote: Make sure to set up your environment variables before starting the application.
- Welcome Page: Learn about features and get started
- Ad Generation:
- Enter your ad text
- Select tone and platform
- Choose output types (text, poster, video)
- Generate your perfect ad
- Download/Copy: Get your generated content
- Login: Use credentials
admin/admin(click "Admin" in top-right corner) - Dashboard Hub: Overview of key metrics and business intelligence
- Recent Activity: Click "Recent Activity" tab to view detailed generation history
- Customer Insights: Click "Customer Insights" tab for comprehensive feedback analysis
- RAG Analytics: Monitor knowledge base performance and generation quality
- Business Intelligence: Track ROI, conversion rates, and platform performance
This project implements an advanced Agentic RAG (Retrieval-Augmented Generation) system with LangGraph orchestration and Feedback-Driven Improvement Loops.
Detailed architecture documentation: architecture.md
- Features & Updates: Detailed feature list and recent enhancements
- Tech Stack & Structure: Technology stack and project organization
- Setup Guide: Complete installation and configuration instructions
- Contributing: Guidelines for contributing to the project
- Content Researcher Agent: Retrieves relevant templates, examples, and guidelines from the knowledge base using semantic search
- Copywriter Agent: Generates platform-optimized ad copy with tone consistency and brand alignment
- Visual Designer Agent: Creates detailed prompts for AI image generation with design specifications
- Video Scriptwriter Agent: Develops structured video scripts with scene descriptions and narration
- Quality Assurance Agent: Validates outputs against quality criteria and provides refinement suggestions
The system implements a sophisticated Pattern Recognition and Improvement Loop that continuously learns and adapts:
Pattern Recognition Engine:
- Aggregates user feedback from MongoDB database
- Identifies positive highlights, improvement suggestions, and keyword patterns
- Calculates average ratings and sentiment trends per platform/tone combination
Adaptive Generation Process:
- Injects feedback insights into agent contexts before generation
- Guides copywriting, visual design, and video scripting based on user preferences
- Continuously improves output quality through iterative learning
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
Detailed contribution guidelines: contributing.md
- React team for the amazing framework
- Lucide for beautiful icons
- Vite for the fast build tool
- The open-source community for inspiration
If you have any questions or need help, please contact shree.xai.dev@gmail.com
Built with ❤️ by Shreeraj Mummidivarapu