Metabase的动态

?? How I Automated My House Hunt with a Telegram Bot, DuckDB, and Metabase ???? Manually checking property listings every day? No thanks. Instead, I built a fully automated system that scrapes new properties, sends them to me via Telegram, and stores them in MotherDuck for analysis. ?? The stack: ? Scraping: Python + BeautifulSoup ? Notifications: Telegram bot ? Storage: MotherDuck (DuckDB in the cloud) ? Data Cleaning: SQLMesh by Tobiko (bye-bye 10M€+ auction glitches) ? Visualization: Metabase dashboard Now, instead of refreshing a website like a caveman, I track trends, filter junk, and get alerts—all hands-free. ?? Next steps? Maybe enriching data with more info, like schools around, supermarkets and things like that or creating alerts for properties matching specific criteria. The possibilities are endless. As someone managing data teams by day, building this project reminded me of a few things: - Hands-on work keeps you sharp: Too many data leaders forget what it’s like to build something from scratch. - Simple systems are maintainable systems: Every time I was tempted to add complexity, I asked, “Will this actually help me buy a property in Italy?” - Personal projects let you experiment fast: I tested MotherDuck and SQLMesh here before considering them for work projects. Want to see how I built it? A post on my blog plus the code and a link to the dashboard are in the first comment! #DataEngineering #Automation #RealEstateTech #DuckDB #Metabase #SQLMesh