Publish LinkedIn posts from RSS using Ollama, PostgreSQL and Telegram approval

This workflow automatically converts incoming RSS/news articles into structured, AI-enriched content records. It uses a local AI model to summarize articles, extract metadata, and classify content before storing it in PostgreSQL. A human-in-the-loop approval step via Telegram ensures only validated content is published to LinkedIn, maintaining quality and brand safety.

⚙️ How it works
Fetch RSS feed articles → AI summarizes content using Ollama → Extracts category, keywords, sentiment, and key points → Stores structured data in PostgreSQL → Sends content to Telegram for human approval → User approves or rejects → Approved content is marked in database → Automatically publishes approved posts to LinkedIn

📦 Requirements

n8n (latest version)
PostgreSQL database
Telegram Bot API credentials
LinkedIn Developer App (OAuth2)
Ollama (local AI runtime) with model like qwen2.5:3b
RSS feed source (e.g., Ars Technica or any news feed)

📊 Database Table Required: rss_feed_articles

Stores AI-processed articles, metadata, and approval status for controlled publishing workflow.