Room Finder Orchestrator

An Apify Actor that orchestrates and unifies real estate scraping from multiple sources. It uses an LLM (via Apify OpenRouter proxy) to process natural language queries into specific scraper inputs. LLM features are automatically enabled when running on the Apify platform using your APIFY_TOKEN.

Features

• 🎯 Unified Interface: One input to rule them all (Facebook, eSpolubydleni, UlovDomov).
• 🧠 AI-Powered: Converts natural language (e.g., "Student room in Prague < 10k") into structured scraper configs.
• 🔄 Normalization: Unifies outputs from different sources into a single schema.
• ⚡ Parallel Execution: Runs multiple scrapers concurrently for faster results.

Supported Sources

1. Facebook Groups (via Apify Facebook Groups Scraper)
2. eSpolubydleni.cz (via custom Actor)
3. UlovDomov.cz (via custom Actor)

Input Parameters

Parameter	Type	Description
naturalLanguageQuery	string	(Recommended) Plain English query (e.g., "Cheap flats in Brno").
enableFacebook	boolean	Toggle Facebook Groups scraping.
enableEspolubydleni	boolean	Toggle eSpolubydleni scraping.
enableUlovdomov	boolean	Toggle UlovDomov scraping.
facebookGroupUrls	array	List of Facebook Group URLs to scrape.

Output Schema

All listings are normalized to a unified format:

{
  "source": "facebook", // or "espolubydleni", "ulovdomov"
  "sourceUrl": "https://...",
  "title": "Sunny room in Vinohrady",
  "priceCzk": 15000,
  "location": "Prague",
  "district": "Vinohrady",
  "description": "...",
  "contactName": "Owner Name",
  "contactPhone": "+420...",
  "scrapedAt": "2024-01-20T12:00:00.000Z"
}

Setup

1. Configure Sub-Actors: Ensure the Actor IDs for the scrapers are set in src/config/constants.ts.
2. Build: npm run build
3. Run: apify run