RAG-Wissen — Production RAG for our own knowledge base
RAG-Wissen is our internal RAG system for advisory knowledge, technical documentation, and market intelligence. Runs in production with Qdrant v1.17.1 and provides the research base for our blog articles.
Python 3.13QdrantFastAPIMCPHarrier-Embeddings
Embedding model choice
We benchmarked five embedding models for German content. Harrier won with +50% over mxbai and +12% over Qwen3. Result: production use since April 2026.
What we learn from it
Embedding model selection is non-trivial — standard recommendations for English content often work poorly for German text. We bring this benchmarking experience into every RAG strategy workshop.