Qdrant
Vector database built for filtered search at scale.
- made by
- Qdrant
- source
- Official Qdrant site
Magic Ship is one shop in Vancouver, BC, working remotely with clients worldwide. We are not a partner, reseller, or certified vendor of Qdrant - we just build with this.
What Qdrant is
Qdrant is an open-source vector database written in Rust, available self-hosted or as a managed cloud. It stores a payload next to each vector and applies filters during index traversal rather than before or after it, supports sparse vectors for hybrid search, and offers quantization to cut memory use.
How we use it
We choose it when filtered vector search is the hard part - per-tenant, per-permission, or per-time-window queries where a naive filter destroys either recall or latency. Self-hosting in the client's own cluster keeps embeddings inside their network. Named vectors let one collection hold two embedding models at once, which is how a migration between models happens without downtime.
Where it is the wrong choice
It is another stateful service to run, back up, and keep in sync with the source of truth. If the corpus fits comfortably in Postgres, adding Qdrant buys complexity rather than recall.
Service lines it turns up in
Related tools
More in Retrieval and vector
Other tools in the same service lines
Building something on Qdrant?
Send the problem rather than a job spec. You get an answer on scope, on fit, and on whetherQdrant is even the right call for it.
Start a projectQdrant and Qdrant are trademarks of their respective owners, used here to say what we work with.