Search
About
The focus of my technical career has been in the intersection of search and knowledge management and how we can make use of this information in practical ways.
Currently I work at Weaviate, where we develop improvements to vector indexes, such as new quantization, graph algorithms, and low-level performance work to make search fast and affordable.
Previously I was a principal consulting architect at Elastic, and before that worked in various data analytics and engineering roles.
I'm based in Brisbane, Australia but originally from New Zealand.
Recent Work
-
2026-09-17
4-bit Rotational Quantization
Rotation-based 4-bit compression for vector indexes in Weaviate 1.39 — SIMD work, a centered tier, scaling analysis, and a TurboQuant comparison.
-
go
go-potion
Static text-embedding inference in pure Go — vectors identical to model2vec, 8–15× faster encoding via a native tokenizer and SIMD.
-
λ
rf
A persistent, Scheme-programmable shell and agent harness — structured pipes, built-in LLM and embedding primitives.
Home Sector
