John Wagster
Senior Software Engineer
John Wagster is a software engineer at Elastic, where he works on improving Elasticsearch.
Preconditioning Vectors: Making Elasticsearch VectorDB Better Binary Quantization work for every vector
Elasticsearch Labs
Preconditioning Vectors: Making Elasticsearch VectorDB Better Binary Quantization work for every vector
Modern quantization techniques can hurt recall when using older models or embeddings that aren’t normally distributed. Learn how preconditioning fixes these vectors through random orthogonal projection, making BBQ more effective and recovering recall.
Low-memory benchmarking in DiskBBQ and HNSW BBQ
Elasticsearch Labs
Low-memory benchmarking in DiskBBQ and HNSW BBQ
Benchmarking Elasticsearch latency, indexing speed, and memory usage for DiskBBQ and HNSW BBQ in low-memory environments.
Introducing a new vector storage format: DiskBBQ
Elasticsearch Labs
Introducing a new vector storage format: DiskBBQ
Introducing DiskBBQ, an alternative to HNSW, and exploring when and why to use it.
RaBitQ binary quantization 101
Elasticsearch Labs
RaBitQ binary quantization 101
Understand the most critical components of RaBitQ binary quantization, how it works and its benefits. This guide also covers the math behind the quantization and examples.