AI-Driven Content Factory

Overview

An intelligent content generation pipeline that transforms topics into structured educational videos with character-based narration and automated quality assurance.

Features

• Curriculum Generation: Breaks topics into 3-5 structured lessons
• Character Management: Database-driven character library with voice profiles
• Script Generation: AI-powered script writing with character dialogues
• TTS Integration: Multi-voice text-to-speech with consistency checking
• Video Assembly: Automated scene generation with captions
• Quality Assurance: Automated checks for audio sync, character consistency, and caption alignment

Setup Instructions

1. Prerequisites

• Python 3.8+
• FFmpeg installed
• API Keys for Groq, Together AI (optional)

2. Installation

# Clone repository
git clone <repository-url>
cd content-factory

# Install dependencies
pip install -r requirements.txt

# Create .env file
echo "GROQ_API_KEY=your_key_here" > .env
echo "TOGETHER_API_KEY=your_key_here" >> .env

3. Run Application

streamlit run app.py

Architecture

Agent-Based System

• Curriculum Agent: Generates structured lesson plans
• Script Agent: Creates character-based video scripts
• QA Agent: Performs quality checks and auto-corrections

Core Modules

• Character Manager: SQLite-based character library
• TTS Manager: Voice generation with consistency
• Video Generator: Scene assembly and rendering

Technology Stack

• Frontend: Streamlit 1.28.0
• LLM: Groq API (Llama 3.1)
• TTS: Edge-TTS 6.1.9
• Video: MoviePy 1.0.3
• Database: SQLite3
• Image Processing: Pillow 10.1.0

Cost Estimation

Per Video (60 seconds)

• LLM API: $0.002
• TTS Generation: $0.005
• Image Generation: $0.05
• Video Rendering: $0.05
• Storage: $0.002
• Total: ~$0.11 per minute

Scalability Considerations

• Batch processing for multiple videos
• Caching for repeated characters
• CDN for video distribution
• Serverless architecture for rendering

Demo Video

[Link to 15-20 minute walkthrough]