Stores and indexes high-dimensional embeddings for scalable ANN vector search. Distributed, Kubernetes-native architecture with multiple index types, GPU acceleration, and hybrid dense+sparse search — suitable for semantic search, RAG, and recommendation pipelines.
