🎬 Transcript-to-Ad Generator

Transform spoken content into high-converting video ads in seconds.

A sophisticated AI-driven pipeline that ingests audio/video transcripts, analyzes them using NLP, and automatically generates ready-to-use video ad scripts, storyboards, and previews.

---

✨ Features

• Deep Content Analysis: Uses NLP (spaCy) to extract keywords, named entities, and key highlights from transcripts.
• Intelligent Ad Generation:
  • LLM Mode: Uses structured prompting to create professional ad copy with compelling CTAs.
  • Simple Mode: Deterministic rule-based generation for fast local testing.
• Visual Storyboarding: Automatically maps ad copy to visual frames.
• Video Previews: Renders MP4 video previews on-the-fly using MoviePy.
• Interactive UI: A polished Streamlit interface with premium aesthetics, dark mode, and real-time feedback.
• Background Processing: Supports asynchronous rendering via Redis Queue (RQ) for heavy workloads.
• Smart Caching: Built-in system for caching expensive API calls and rendering steps.
• Production Ready: Dockerized and includes robust logging and error handling.

🛠️ Technology Stack

• Frontend: Streamlit (Custom CSS for premium UI)
• Core Logic: Python 3.9+
• NLP: spaCy, TextBlob
• Video Processing: MoviePy, Pillow
• Async Workers: Redis, RQ (Redis Queue)
• Containerization: Docker & Docker Compose
• Quality: pytest, flake8, mypy, pre-commit

🚀 Getting Started

Prerequisites

• Python 3.9+
• FFmpeg (required for local video rendering)
• Docker (optional, for containerized execution)

📦 Quick Start (Local)

1. Clone the repository:

   git clone https://github.com/your-org/transcript-to-ad.git
   cd transcript-to-ad

2. Set up environment:

   python -m venv .venv
   # Windows:
   .\.venv\Scripts\Activate
   # Linux/Mac:
   source .venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Run the application:

   streamlit run app.py

5. Open http://localhost:8501 in your browser.

🐳 Run with Docker (Recommended)

Easily spin up the entire stack including the Redis worker:

docker-compose up --build

The app will be available at http://localhost:8501.

📂 Project Structure

transcript-to-ad/
├── core/                   # Business Logic & Pipeline
│   ├── analysis.py         # NLP & Text Analysis
│   ├── ad_generator.py     # Ad Copy Generation Logic
│   ├── video.py            # Video Rendering (MoviePy)
│   ├── pipeline.py         # Main Orchestration
│   ├── gap_scanner.py      # Opportunity Detection
│   ├── llm.py              # LLM Integration
│   ├── nlg.py              # Natural Language Generation
│   └── queue.py            # Redis Queue Interface
├── scripts/                # Utility Scripts (Worker, Redis helpers)
├── tests/                  # Unit & Integration Tests
├── app.py                  # Main Streamlit Application
├── run_local.py            # CLI Runner for Pipeline
├── Dockerfile              # App Container Config
├── Dockerfile.worker       # Worker Container Config
└── requirements.txt        # Python Dependencies

🧪 Testing & Quality

Run the test suite to ensure everything is working correctly:

# Run unit tests
pytest tests/

# Run type checks
mypy core/

# Run linter
flake8 core/

⌨️ CLI Usage

You can also run the pipeline directly from the command line without the UI:

python run_local.py

This will process a sample transcript and print the detailed analysis and ad copy to the console.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

---

Generated for the Transcript-to-Ad Project.