vector search

1. October 27, 202631 min read

   Storing Quantized Embeddings in Apache Iceberg: Scalar 8-Bit Codes vs Product Quantization

   How SQ8 and product quantization shrink Iceberg embedding columns 4x to 64x, how to store codes and codebooks, and what breaks.

2. September 21, 202631 min read

   Vector Search Directly Over Iceberg Tables, and When You Still Need a Vector Database

   Embeddings are just columns. When exact vector search over Iceberg scans beats a vector database, when it does not, and how to lay out tables.

Browse all posts