- Title: Kyiv Regional News Parser & AI Rewrite Pipeline
- Category: Media Automation & Web Scraping
- Context: Regional city news portal (kyiv1.com) requiring rapid coverage of emergency and municipal news from official public sources (State Emergency Service, Kyiv City Administration).
- Problem: Manual monitoring of websites and Telegram channels was time-consuming. Copying official bureaucratic text ("officialese") created duplicate content and delayed emergency news delivery.
- Solution: Built an automated scraping pipeline in n8n that ingests RSS/HTML announcements, fetches full article bodies, filters them by geographical relevance using AI, rewrites the content into engaging journalistic style, and generates WordPress drafts automatically.
- Process:
- Scraping & Parsing: Configured HTTP Requests and HTML CSS selectors to extract full article bodies from sources without RSS feeds.
- AI Processing: Created prompts to convert official reports into structured news (Headline, Lead, Body, SEO Tags, Meta Description) returning strict JSON.
- WordPress Integration: Automatic article draft creation via WordPress REST API with duplicate URL checks in the database.
- Results:
- 📈 +300% publishing speed: Breaking news is drafted and ready within 2 minutes of source release.
- 💰 Resource Efficiency: Automates the routine work of an entire news desk shift.
- 🤖 SEO Friendly: Preserves accurate facts and dates while producing unique content.
- Tech Stack: n8n, OpenAI API (GPT-4o-mini), HTML Extract Node, WordPress REST API, JavaScript.
- Visuals:


