Topic

#Vector Search

7 articles on this topic.

Qdrant SQL with QQL 0.4.0: AST parameter binding, whole-point bulk upsert, one-roundtrip BATCH blocks, REST conversion, and checkpointed cluster migration

Qdrant SQL with QQL 0.4.0: Bulk Upsert, Batch & Migrate

Build fast Qdrant SQL applications with QQL 0.4.0. Get AST parameter binding, bulk upsert, one-roundtrip BATCH blocks, and cluster migration with resume.

Sep 12, 202620 min readRead

System architecture of GeoSmart showing upstream DuckDB spatial enrichment and Qdrant multi-variable continuous decay scoring

QdrantVector SearchGeospatial

Qdrant Geospatial Search: Fix the Boundary Cliff with Decay

We replaced Qdrant geo_radius hard filters with GaussDecay and ExpDecay on 8,317 Berlin stays. Ring loss 37% to 0% in 8.1 ms. Code, data, and the 10-persona eval are on GitHub.

Sep 3, 202611 min readRead

Diagram of JurisBoost single query_points call with RRF prefetch, ColBERT rescore, and statutory FormulaQuery

QdrantVector SearchRAG

Qdrant Hybrid Search for Legal RAG Tested on 2,034 Cases

We ran Qdrant hybrid search (RRF + ColBERT + FormulaQuery) on 2,034 Supreme Court precedents. Top-1 went from 31.8% to 86.4%. Code, data, and the 22-query eval are on GitHub.

Sep 2, 202616 min readRead

QQL 0.3.0 cover: bound query with :q and :tenant, WHERE isolation, idf = WHERE scoring, and matching BM25

QQLQdrantBM25

QQL 0.3.0: Bound Queries, Tenant IDF, Matching BM25

Pin QQL 0.3.0 (language 1.5). Bind :name params, write idf = WHERE for tenant BM25, mix local and server sparse vectors. Re-embed old collections.

Aug 30, 202610 min readRead

Editorial diagram of a shared Qdrant collection mixing twelve tenant vocabularies into global IDF, then scoping params.idf.corpus to one tenant

QdrantBM25Vector Search

Qdrant 1.19 Per-Tenant IDF: Ranking Changes, Not nDCG

We ran Qdrant 1.19 per-tenant IDF on 457k CQADupStack docs. Top-1 changed on 20% of queries. Labeled nDCG barely moved. Methodology, numbers, and the API.

Aug 22, 202615 min readRead

Cream editorial cover: Local Vector Search with in-process FastEmbed and qdrant-edge, no remote server

QQLQdrantPython

Run Local Vector Search in Python Without a Qdrant Server

Build local vector search in Python and Node with pyqql-edge. Embed Qdrant and FastEmbed in-process with zero server setup, disk persistence, and real scores.

Jul 30, 202612 min readRead

Cream editorial cover: QQL WASM playground browser split between local analyze() and Qdrant REST Execute

QdrantWebAssemblyQQL

Inspect Qdrant Queries in the Browser Before They Hit the API

QQL WASM parses and projects REST routes locally while you type. Execute only when you click. No cluster round-trip in the editor.

Jul 30, 202611 min readRead

All posts