Production guides

The language pages explain individual constructs. These guides combine them into production retrieval patterns and architectural solutions.

Side-by-side comparison of declarative QQL queries vs verbose nested REST JSON and Python payloads.

Build dense and sparse candidate stages, then fuse them with RRF or DBSF.

Keep isolation, shard routing, index layout, and partition lifecycle distinct.

Enforce tenant and authorization predicates at the AST level with inject_filter.

Run local vector search and ONNX FastEmbed models in Python and Node.js.

Copy schema and points between clusters with dry-run, resume, and cutover.

Export a collection to a replayable .qql script and restore it.

Resolve vector roles and choose late-interaction or cross-encoder reranking.

Choose between Qdrant REST, gRPC, and in-process edge execution.

Typed shard keys, strict scripts, canonical statement strings, and SDK behavior changes.