Knowledge & Data · Software component
Approximate Vector Search Retriever
Software componentKnowledge & DataKnowledge & Dataarc:ApproximateVectorSearchRetriever
A vector retriever that uses approximate nearest-neighbour search (e.g., HNSW, IVF), trading exactness and determinism for sub-linear query time.
Responsibility. Retrieves approximately nearest vectors with sub-linear latency.
Also known as: ANN search
Variant of Vector Retriever abstract
When to choose. Choose when sub-linear, millisecond-latency search over millions of vectors matters more than exact, reproducible top-K results.
Relationships
is configured by structural
reads dependency
- Vector Index Store abstract Ch6.2A Ch6.5
is triggered by dynamic
is evaluated by assurance
alternative to variability
Design guidance
- MAY use deterministic seeding to make approximate search repeatable at the cost of occasionally missing optimal results.
- SHOULD be adopted at production scale, accepting quality-speed trade-offs, when exact scans take seconds.
Quantitative guidance
As stated by the sources; verify before use.
- ANN returns results in 10-100 ms at 95-99% of exact-search accuracy; tunable down to ~90% for speed (Ch6.2A).
Classification
- Patterns
- HNSWIVFDeterministic seeding for repeatable ANNAnnoyDiskANNHierarchical greedy graph search (O(log N))Adaptive ef: retry at higher ef when confidence is low
- Quality attributes
- Performance efficiency (ISO/IEC 25010)
Sources
- Ch5.8: T. Nguyen, "Semantic Memory," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.8. ISBN: 9798244538229.
- Ch6.2A: T. Nguyen, "Vector Database Selection," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.2A. ISBN: 9798244538229.
- Ch6.5: T. Nguyen, "Production RAG Systems," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.5. ISBN: 9798244538229.