A self-hosted real estate crawler that monitors German listing platforms and sends instant notifications for new properties.

Fredy is a self-hosted web application that monitors German real estate platforms for new property listings. It automates the process of searching for apartments and houses by scraping multiple websites and sending instant notifications to the user when new results match their criteria. This allows users to react quickly to new listings in competitive housing markets.
The software is deployed as a server application and can be run via Docker, Node.js, or through the Unraid community store. It includes a dedicated web interface for managing search jobs and configuring notification channels. Users provide search URLs from supported platforms, and the system monitors these pages at customizable intervals to identify new entries.
The application architecture relies on three core concepts: providers, which are the real estate platforms; adapters, which are the notification channels; and jobs, which link specific providers to specific adapters. To avoid detection, the system utilizes a hardened headless browser that mimics real user fingerprints. For users deploying on a VPS, the software supports SOCKS5 residential proxies to avoid IP reputation blocks from datacenter ranges.
For advanced data interaction, the built-in MCP Server allows the application to connect to Large Language Models such as Claude or ChatGPT. This enables users to query their collected real estate data using natural language or use the LLM to enrich existing listings by checking them online. This makes the tool suitable for users who want to organize and analyze their property search data programmatically.
Fredy is designed for individuals searching for housing in the German market who require immediate alerts for new listings.
Join our newsletter to get shiny new open source software delivered to your inbox. Unsubscribe anytime.