Case Study: Popety.io automates Swiss zoning data extraction with Astronomer’s Astro

Popety.io automates 40,000+ hours of zoning extraction with Astronomer

Popety.io, a Swiss PropTech company, faced the challenge of manually extracting complex building regulations from thousands of fragmented documents across Switzerland's cantons and communes. This slow, unscalable process prevented them from providing customers with quick answers on property development rights. They turned to Astronomer, using its managed Apache Airflow service, Astro, to orchestrate their existing data pipelines, but needed a solution to automate the extraction of this unstructured data.

Astronomer's Astro platform provided the orchestration foundation for Popety.io to build a fully automated RAG pipeline. This solution uses LLMs to read, retrieve, and extract structured zoning data from scanned PDFs, which is then fed directly into their product. The results were transformative, automating over 40,000 hours of manual work. The pipeline now runs over 7,000 LLM extraction tasks monthly on Astro, providing customers with instant, traceable answers on what can be built on any parcel.

Popety.io