About

I build and scale storage and distributed systems infrastructure.

I’m currently a Tech Lead in Google Cloud Storage, where I own the offline data plane behind Blobstore — the storage layer under YouTube, Google Workspace and GCS — and I’m responsible for the durability, byte efficiency and reliability of hundreds of exabytes of data. Before that I led reliability for MindMeld, a distributed in-memory filesystem built on RDMA that serves BigQuery and AlphaFold2.

Previously: Tech Lead for infrastructure at Canva, where I built the SRE organisation from scratch and set technical direction across 160 engineers; Production Engineer on ZippyDB at Facebook, the key/value store behind Ads, Feed, Instagram and WhatsApp; and SRE Tech Lead at Dropbox.

I started out as a sysadmin in Perth, Western Australia, fixing Outlook crashes and writing VBScript. The oldest posts here are from that era and I’ve kept them on purpose.

I write about systems software, performance and reliability. Opinions are my own.

• github.com/cavaliercoder — my open source projects, and the archive
• github.com/cavaliergopher — my maintained Go packages, including grab