How Poliscope Made All of Germany's Local Politics Searchable on One Server with Qdrant
Blog post from Qdrant
Poliscope aggregates agendas, minutes, proposals, and planning documents from roughly 4,800 German municipal council portals, making local political activity across about 90% of municipalities searchable through a single platform. After processing 4.3 TB of PDFs with OCR, the company built a self-hosted hybrid retrieval system using Qdrant that indexes 130 million points, including 117 million document chunks, and combines dense semantic search with BM25 keyword matching, geographic filters, result grouping, and reranking. Driven by data-sovereignty requirements and a preference for operating within Germany, Poliscope runs Qdrant as a single open-source node on a German bare-metal server rather than using a managed cloud service. Performance improvements including state-based tenant partitioning, prefix indexes for hierarchical regional IDs, and TurboQuant vector compression reduced typical end-to-end search time to an average of 0.87 seconds while cutting storage use and maintaining retrieval quality. The small team plans to extend its archive from 2024 back to 2016 and eventually add a second node as the corpus expands.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| MCP | 3 | 11,903 | 1,228 | 233 | +12% |
| Vector Search | 3 | 1,977 | 414 | 134 | +18% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.