AI Content Generation Engine — Automated Blog Publishing System

Overview

AI Content Generation Engine automates a full research-and-write loop for a blog. It finds what is trending, researches it, writes an original article and publishes it to WordPress.

How it works

• Discovery: popular Pinboard bookmarks are scraped and parsed from HTML, then GPT-4o-mini filters them down to AI-related topics.
• Summarization: each relevant link is passed to a reusable scrape-and-summarize sub-workflow that downloads the page, converts it to Markdown and condenses it.
• Ideation: the summaries act as inspiration for an LLM that proposes a unique article topic.
• Research: a Google deep-research sub-workflow (Google Custom Search) generates multiple queries, fetches results and keeps only content the AI judges relevant.
• Writing and publishing: the article is written as structured long-form content, converted from Markdown to HTML and created as a WordPress post.

Engineering notes

The system is designed around sub-workflows called through Execute Workflow nodes, so the scraping and research components can be reused by other automations. Seven LLM chains with structured output parsers keep each stage's output strictly typed, and filter nodes drop broken pages and irrelevant results before they reach the writer.