Qdrant

Qdrant is a vector search company founded in 2021 and headquartered in Berlin, Germany. It was created to solve the problem of finding similar items among massive collections of unstructured data such as text, images and audio, a task traditional databases handle poorly.

The product is an open-source vector search engine, written in Rust, that indexes high-dimensional embeddings and returns nearest-neighbor matches with real-time metadata filtering. It is offered as self-hosted software, as fully managed Qdrant Cloud on AWS, GCP and Azure, or as Qdrant Hybrid Cloud for customer-controlled Kubernetes; customers include TripAdvisor, HubSpot, Canva and Deutsche Telekom, using it for semantic search, recommendation and retrieval-augmented generation.

1. Vector search engine: Rust-based approximate nearest-neighbor indexing over high-dimensional embeddings with real-time metadata filtering.

2. Managed cloud: Qdrant Cloud runs fully managed clusters on AWS, GCP and Azure.

3. Hybrid deployment: Qdrant Hybrid Cloud lets customers operate managed clusters inside their own Kubernetes environment.

4. Cloud inference: built-in embedding generation removes the need for a separate inference service.

5. Application use cases: supports retrieval-augmented generation, semantic search, recommendation and anomaly detection.

6. Open-source core: the engine is open source with an active developer community and enterprise customers such as HubSpot and Canva.