AI Content Writer

A Streamlit application that generates content for blog posts, LinkedIn, and X (Twitter) based on a given subject and target audience. The application uses AI to create tailored content for different platforms in parallel.

This project is a Python implementation using LangChain of a simple n8n workflow for content creation. For visual demonstrations of the workflow, check out these YouTube videos:

• English: AI Automation by Nate Herk
• German: KI Automatisierung by Sebastian Claes

🔍 Prerequisites

Docker Desktop

This application requires Docker Desktop to run. If you don't have Docker Desktop installed, follow these steps:

Windows

1. Download Docker Desktop from Docker's official website
2. Run the installer and follow the installation wizard
3. Start Docker Desktop from the Start menu
4. Wait for Docker to start (you'll see the Docker icon in the system tray turn solid)

macOS

1. Download Docker Desktop from Docker's official website
2. Open the downloaded .dmg file and drag Docker to your Applications folder
3. Start Docker from your Applications folder
4. Wait for Docker to start (you'll see the Docker icon in the menu bar)

Linux

1. Follow the instructions for your specific distribution on Docker's documentation
2. Start Docker with sudo systemctl start docker
3. Verify installation with docker --version

🚀 Installation and Setup

1. Clone the repository:

   git clone https://github.com/ankoehn/ai-content-writer.git
   cd ai-content-writer

2. Create your environment file:

   cp .env_template .env

3. Edit the .env file and add your API keys:

   • Required: TAVILY_API_KEY for search functionality
   • Required: OPENAI_API_KEY for OpenAI models (default provider)
   • Optional: DEEPSEEK_API_KEY if using DeepSeek models

4. Configure your LLM provider in the .env file:

   • For OpenAI (default):

     LLM_PROVIDER=openai
     LLM_MODEL=gpt-4o
     

   • For DeepSeek:

     LLM_PROVIDER=deepseek
     LLM_MODEL=deepseek-chat
     DEEPSEEK_API_KEY=your_deepseek_api_key
     

🏃‍♂️ Running the Application

Start the application using Docker Compose:

# Run in foreground
docker compose up

# Or run in background
docker compose up -d

Access the application in your web browser at:

http://localhost:8085

📝 Usage

1. Open your browser and navigate to http://localhost:8085

2. Fill in the form with:

   • Campaign name
   • Content subject (what you want to write about)
   • Target audience (who the content is for)

3. Click "Create" to generate content

4. View the generated content for Blog, LinkedIn, and X

5. Access previously generated content from the history sidebar

6. Export your content history to Excel using the export button

⚙️ How It Works

1. The application uses the Tavily search engine to find relevant information about the content subject
2. It then processes this information using three different AI agents specialized for each platform:
   • Blog Agent: Creates a concise two-paragraph blog post
   • LinkedIn Agent: Creates an engaging LinkedIn post with emojis and hashtags
   • X Agent: Creates a short, impactful tweet with emojis and hashtags
3. Generated content is saved to a local JSON file for persistence

📁 Project Structure

• app.py: Main Streamlit application
• docker-compose.yml: Docker Compose configuration
• Dockerfile: Docker container configuration
• requirements.txt: Python dependencies
• writer/: Core content generation functionality
  • ai/: AI agents and LLM processing
  • searchengine/: Search engine integration
  • model.py: Data models
  • config.py: Configuration settings
  • utils/: Utility functions

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.