AI Content Engine

The AI Content Engine is a high-performance system designed to automate the creation of short-form video content (Reels, TikToks) and cross-platform social media assets. It leverages advanced Large Language Models (LLMs) and generative media APIs to produce professional-grade videos from a single topic input.

Features

• Automated Research: Scrapes trending topics and audience sentiment from social platforms.
• Multimodal Generation:
  • Scripting: Utilizes Gemini 2.x/3.x for viral script development and LinkedIn post generation.
  • Voiceover: High-fidelity narration via SiliconFlow (Fish Audio) with a robust gTTS fallback.
  • Visuals: Dynamic video generation using FLUX/Wan2.1 models with automatic image/gradient fallbacks.
• High-End Rendering: Orchestrated via MoviePy with custom ImageMagick configurations for text overlay and cinematic effects.
• Automated Delivery: Direct integration with Google Drive for asset storage and distribution.
• Active Development: Regularly updated with new AI model integrations and pipeline optimizations.

System Architecture

The engine follows a modular service-oriented architecture:

1. Core Configuration: Centralized settings management using Pydantic Settings and strict environment validation.
2. Scraper Service: Handles topic discovery and data enrichment.
3. AIBrain Service: Manages LLM orchestration, including robust JSON parsing and model failover logic.
4. Voice Service: Connects to TTS providers with automatic error handling and local synthesis fallback.
5. Visual Service: Generates background media (video/images) tailored to the script's emotional context.
6. Video Engine: Performs the final assembly, synchronizing audio, visuals, and dynamic subtitles.

Setup Requirements

External Dependencies

1. Python 3.10+: Recommended environment.
2. ImageMagick: Required by MoviePy for text rendering. Ensure the binary path is correctly set in core/config.py or your environment variables.
3. FFmpeg: Required for video transcoding and audio merging.

Installation

# Clone the repository
git clone https://github.com/Sameershahh/ai-content-engine.git
cd ai-content-engine

# Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

Configuration

Create a .env file in the root directory based on .env.example:

1. Copy the example .env file:

   cp .env.example .env

2. Configure your API keys and parameters:

   GEMINI_API_KEY=your_gemini_key
   SILICONFLOW_API_KEY=your_siliconflow_key
   DRIVE_CREDENTIALS_PATH=credentials.json

3. Place your Google Service Account JSON key in the root directory as credentials.json. Ensure the service account email is added as an Editor to your target Google Drive folder.

Usage

To generate a new reel and its associated metadata:

# Generate content for a specific topic
python run_pipeline.py "The impact of quantum computing on cybersecurity"

# Direct run with default trending discovery
python run_pipeline.py

Outputs will be generated in the outputs/ directory and automatically uploaded to the configured Google Drive folder.

Failover Mechanisms

The system is designed for high availability:

• LLM Failover: If the primary Gemini model is unavailable, the system automatically cycles through fallback models.
• TTS Fallback: If the SiliconFlow API returns a 401 or connection error, the system seamlessly switches to gTTS to ensure the pipeline completes.
• Visual Fallback: If video generation fails, the system transitions to static image generation, and finally to cinematic gradient rendering.

License & Disclaimer

This project is specialized for professional content automation. Ensure compliance with API provider terms of service (Google, SiliconFlow) and platform guidelines (Reddit, LinkedIn) before deployment.

---

Author
Sameer Shah — AI & Full-Stack Developer
Portfolio