/azlo-blog-generator

An AI-powered content pipeline that automates the creation of SEO-friendly Hugo blog posts using Python, Docker, and Google's Gemini AI.

Primary LanguagePython

๐Ÿค– AI-Powered Blog Generation System

Typing SVG

Built by Azlo.pro

GitHub stars License Python


๐Ÿš€ Why This Exists (And Why I'm Sharing It)

As a freelance developer who's built custom AI integrations for businesses, I kept running into the same problem: companies needed consistent, brand-aligned content but couldn't afford the time or resources to produce it manually.

This started as an internal tool for my own content strategy. After seeing 10x improvements in my content output quality and consistency, I realized other developers and businesses could benefit from the same approach.

๐Ÿ’ก I'm releasing this freely because I believe in building in public and giving back to the community that's taught me so much.

โœจ What Makes This Different

๐Ÿง  Smart Context Integration

Unlike generic AI tools, this system:

  • Understands your business - Injects brand voice into every piece
  • Maintains consistency - Strategic positioning across all content
  • Learns your style - Adapts to your unique requirements
AI Brain

๐Ÿ“ฆ Complete Production Pipeline

Creates everything you need:

  • Article content - SEO-optimized, publication-ready
  • Visual assets - Context-aware AI-generated images
  • Meta data - Hugo front matter, tags, categories
  • Ready to deploy - One click to your static site

๐Ÿ“Š Real-World Results

Metric Before After Improvement
โฑ๏ธ Time per Article 4-6 hours 15 minutes 95% reduction
๐ŸŽฏ Brand Consistency Variable 100% Perfect
๐Ÿ“ˆ Content Volume 2 posts/week 20 posts/week 10x increase
๐Ÿ’ฐ Cost per Article $200 $5 97% savings
"The difference between this and ChatGPT is like the difference between a custom-built race car and a rental sedan." Both get you places, but only one is built for performance.

๐Ÿ›  Technical Architecture

graph TB
    A[๐Ÿ“ CSV Input] --> B[๐Ÿค– AI Strategy Engine]
    B --> C[๐Ÿ“‹ Content Planning]
    C --> D[โœ๏ธ Article Generation]
    C --> E[๐ŸŽจ Image Creation]
    D --> F[๐Ÿ“ฆ Hugo Bundle]
    E --> F
    F --> G[๐Ÿš€ Deploy Ready]
    
    style A fill:#6366f1,stroke:#333,color:#fff
    style B fill:#10b981,stroke:#333,color:#fff
    style G fill:#f59e0b,stroke:#333,color:#fff
Loading

๐ŸŽฏ AI Models & Flexibility

Google Gemini OpenAI GPT-4 Anthropic Claude

โšก Infrastructure

AWS Google Cloud Python

โš™๏ธ Getting Started

Ready?

Prerequisites

You'll need:

  • ๐Ÿ Python 3.8+ (the language that powers this)
  • ๐Ÿ”‘ Google Cloud Project with Vertex AI enabled
  • ๐ŸŽฏ Gemini API key for configuration generation
  • โฑ๏ธ 15 minutes to set everything up

๐Ÿš€ Quick Setup

๐Ÿ“ฆ 1. Clone and prepare
git clone https://github.com/nibbabob/azlo-blog-generator.git
cd azlo-blog-generator
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
๐Ÿ” 2. Configure your API access
export GEMINI_API_KEY='your-api-key-here'
gcloud auth application-default login  # For GCP integration
โš™๏ธ 3. Customize for your business

Open config.yaml and make it yours:

  • GCP settings: Your project ID and preferred region
  • Business context: This is the secret sauce ๐Ÿ”ฅ
  • Output paths: Where you want your Hugo content to live
  • Model preferences: Stick with defaults or choose your preferred AI models
๐Ÿ“ 4. Add your content ideas

Edit blog_titles.csv with your post titles:

title
"Why Custom Automation Beats Generic Tools Every Time"
"The Hidden Costs of Manual Data Processing"
"Building AI That Actually Understands Your Business"
๐ŸŽฌ 5. Generate your content
python -m pip install -r requirements.txt
$env:GEMINI_API_KEY="your-key"
python main.py

Sit back and watch as each title becomes a complete, publication-ready blog post! โœจ

๐ŸŽฏ How It Actually Works

sequenceDiagram
    participant U as User
    participant S as System
    participant AI as AI Engine
    participant G as Content Generator
    participant I as Image Creator
    
    U->>S: Upload CSV with titles
    S->>AI: Analyze business context
    AI->>S: Generate content strategy
    S->>G: Create article content
    S->>I: Generate relevant images
    G->>S: Return optimized article
    I->>S: Return styled images
    S->>U: Deliver Hugo-ready bundle
    
    Note over S,AI: Strategic Planning
    Note over G,I: Content Creation
    Note over S,U: Ready to Publish
Loading

This isn't magic - it's systematic intelligence:

1๏ธโƒฃ Strategic Planning

For each title, the system first generates a comprehensive content strategy:

  • ๐ŸŽฏ Target audience analysis
  • ๐Ÿ” Keyword research and SEO positioning
  • ๐ŸŽจ Brand alignment and messaging strategy
  • ๐Ÿ“ธ Visual style guidelines

2๏ธโƒฃ Content Architecture

Creates a detailed outline that includes:

  • ๐Ÿ“‹ Section-by-section talking points
  • ๐Ÿ–ผ๏ธ Image placement and descriptions
  • ๐Ÿ“ž Call-to-action positioning
  • ๐Ÿ”— Internal linking opportunities

3๏ธโƒฃ Content Generation

Drafts the complete article following the strategic plan:

  • ๐Ÿ—ฃ๏ธ Brand-consistent voice and tone
  • ๐Ÿ” SEO-optimized structure and keywords
  • ๐ŸŽฏ Natural integration of your business positioning
  • โœจ Professional, publication-ready quality

4๏ธโƒฃ Visual Assets

Generates relevant, high-quality images:

  • ๐ŸŒŸ Featured image for social sharing
  • ๐Ÿ“Š In-content visuals that enhance understanding
  • ๐ŸŽจ Consistent style matching your brand guidelines
  • โšก Automatic optimization for web performance

5๏ธโƒฃ Hugo Integration

Assembles everything into publication-ready bundles:

  • ๐Ÿ“„ Proper front matter with metadata
  • ๐Ÿ–ผ๏ธ Optimized image placement and sizing
  • ๐Ÿ” SEO tags and social sharing setup
  • ๐Ÿš€ Ready to deploy to your static site

๐Ÿ”ง Customization & Extensibility

Models Prompts Context

๐Ÿค– Want Different AI Models?

The system is built to be model-agnostic. Currently optimized for Google's ecosystem, but I've architected it to easily support:

๐ŸŸข Currently Supported ๐Ÿ”ต Coming Soon
  • Google (Gemini, Imagen)
  • Anthropic (Claude)
  • OpenAI (GPT-4, DALL-E)
  • Azure OpenAI (Enterprise GPT)
  • Open-source models (Llama, Mistral)
  • Custom fine-tuned models

Need support for your preferred model? ๐Ÿค Reach out - I'm always happy to extend the system based on real user needs.

๐ŸŽฏ Prompt Engineering

All AI prompts are in prompts.py, separated from business logic. This means you can:

  • ๐ŸŽจ Adjust tone and style without touching core code
  • ๐Ÿงช A/B test different content approaches
  • ๐Ÿญ Customize for specific industries or use cases
  • ๐ŸŽ›๏ธ Fine-tune based on your brand guidelines

๐Ÿข Business Context Integration

The azlo_pro_context section in config.yaml is where the magic happens. This is where you define:

  • ๐ŸŽฏ Your company's unique value proposition
  • ๐Ÿ‘ฅ Target audience characteristics
  • ๐Ÿ† Industry positioning and competitive advantages
  • ๐Ÿ—ฃ๏ธ Brand voice and messaging guidelines

๐Ÿ’ก Pro tip: The more specific and detailed your business context, the better your generated content will be.

๐Ÿ’ญ Real Talk: What This System Won't Do

REALITY CHECK
  • โŒ It won't replace strategic thinking. You still need to decide what topics matter to your audience.

  • โŒ It won't replace human oversight. Always review and edit generated content before publishing.

  • โŒ It won't work miracles with poor inputs. Garbage in, garbage out - invest time in good prompts and business context.

  • โŒ It won't be perfect out of the box. Plan to iterate on your configuration based on initial results.

๐Ÿค Community & Support

๐Ÿ”„ Contributing

Found a bug? Have an idea for improvement? Want to add support for a new model?

PRs Welcome Issues Open Discussions Active

I welcome contributions from developers who understand that good automation solves real problems. Check out the issues tab or submit a PR.

๐Ÿข Need Custom Implementation?

If you're looking for:

  • ๐Ÿ”Œ Custom model integrations
  • ๐Ÿข Enterprise deployment support
  • ๐Ÿญ Specialized industry adaptations
  • ๐Ÿ› ๏ธ Full-service content automation setup

I provide custom AI integration services for businesses that need more than an off-the-shelf solution.

๐Ÿ’ฌ Questions or Ideas?

Type Channel
๐Ÿ› Quick questions Open an issue here
๐Ÿค Collaboration Contact me directly
๐Ÿ‘€ See it in action Check out my blog

๐Ÿ“„ License & Usage

Open Source & Commercial Friendly Use it โ€ข Modify it โ€ข Learn from it Attribution appreciated but not required

This project is open source because I believe good tools should be accessible. Use it, modify it, learn from it.

โœ… Commercial use is welcome - if this helps your business grow, that makes me happy.

โค๏ธ Attribution appreciated but not required - though if you build something cool with it, I'd love to hear about it!


๐ŸŒŸ Final Thoughts

From Quality vs Quantity Dilemma To Quality AND Quantity Solution

I built this because I was tired of choosing between content quality and content quantity. As a developer who understands both the technical and business sides of content creation, I knew there had to be a better way.

This system represents hundreds of hours of prompt engineering, error handling, and real-world testing. I'm sharing it because I believe in building in public and helping other developers solve similar problems.

It's not perfect, but it's production-ready. More importantly, it's designed to evolve with your needs and the rapidly changing AI landscape.

If you find value in this work or build something interesting with it, I'd love to connect. And if you need help implementing custom AI solutions for your business, you know where to find me.

Happy automating! ๐Ÿš€


Christian Nielsen Azlo.pro - Building smarter, one line of code at a time.

Built with โค๏ธ and a healthy respect for the power of well-engineered automation.

Website AI Services Contact