HNSW
AssessTechniques
A graph-based approximate nearest-neighbor indexing method for fast vector search.
Why it's here
Placed in Assess: 4 article(s) of evidence from 3 source(s), led by research-stage coverage, with 1 in the last 30 days. Confidence 56%.
Evidence (4)
- 6InfoQ·9/16/2026researchPinterest Upgrades Manas with Quantization for More Efficient Search
Pinterest Engineering has improved its Manas search platform to handle large-scale data more efficiently for search and discovery. The system uses Scalar and Product Quantization to reduce memory use while preserving high recall, and it is moving toward multi-vector models with SSD-backed storage for better relevance matching.
- 5The New Stack·8/23/2026researchWhy real-time AI at scale is difficult
The article argues that real-time AI systems often fail in production because of data pipeline issues rather than model quality. It highlights three common problems: tail latency under heavy concurrent load, stale features hurting accuracy, and vector indexes degrading over time without maintenance.
- 4Hacker News·8/20/2026product_launchSpacetimeDB 2.0 draws scrutiny in technical review
A Hacker News-linked technical review examines SpacetimeDB 2.0 and argues that its promotional benchmarks are not fair or technically sound. The review says the product may have interesting ideas, but its comparisons against other databases are misleading because they pit different trade-offs against each other.
- 6InfoQ·7/15/2026researchUsing PostgreSQL as the Foundation for Enterprise AI Agents
Gwen Shapira describes how teams are using PostgreSQL to support production AI features in mission-critical applications. The presentation highlights multi-modal patterns such as JSONB parsing, HNSW vector indexing, vector quantization, and agent memory management to provide both deterministic and semantic context to LLMs.