AutoAgents Qdrant

Vector store index integration for Qdrant. This adapter works with AutoAgents embedding providers and supports dense retrieval as well as Qdrant's hybrid queries.

Named vectors

autoagents-qdrant supports Qdrant named vectors via the vector-store API:

• Use VectorStoreIndex::insert_documents_with_named_vectors(...) to upsert points with multiple named vector spaces.
• Use QdrantVectorStore::delete_collection_if_exists() when you need an explicit collection reset before first upsert defines dimensions.
• Use VectorSearchRequest::builder().query_vector_name("symbol") to select the vector space at query time.
• Keep omitting query_vector_name (or use "default") for backward-compatible single-vector behavior.