Home / Filtering
Filtering guides
10 published guides. Start with Metadata filtering.
How to combine vector similarity with structured filters like product, region or date, and the trade-offs.
Read the guideLearning path
All guides
Understand it1 guides
Compare and choose3 guides
ComparisonPre-filtering vs post-filteringFiltering before or after the vector search changes recall and speed. How each works and which to use.ComparisonVector databases with metadata filteringVector databases compared on filter types, pre-filter support and speed under heavy filtering.ComparisonSearching vs filteringSearch ranks by relevance, filters include or exclude. Why mixing them up causes empty or bad results.
How it works5 guides
ExplainerFiltered ANN searchWhy restrictive filters break approximate search and the index designs (like ACORN) that fix it.ExplainerFilter selectivityHow the share of documents a filter keeps decides which execution plan is fastest.ExplainerBitmap filtersHow bitmaps make structured filtering fast and how vector indexes use them.ExplainerPermission-aware searchMaking sure search only returns what each user may see, with document-level ACLs enforced at query time.ExplainerMulti-tenant semantic searchIndex-per-tenant versus shared index with tenant filters, and how each scales and isolates data.