William Ho

At Etsy, I architect and build Rust retrieval services powering search for millions of daily queries at under 100ms latency. I also develop Scala and Apache Beam pipelines that generate embeddings for more than 100 million listings and millions of users each day.

I partner with ML engineers and product teams on retrieval models and personalization, contributing to a 3% lift in search-driven purchases through A/B testing. My work spans Google BigTable, BigQuery, GCP Dataflow, Kubernetes, and automated Bazel-based CI/CD.

Previously, I built Rust feature-store and data-lineage services at StandardAI, and spent nearly seven years scaling Scala microservices, ETL pipelines, subscription workflows, and streaming systems at iHeartRadio. I introduced Rust for performance-critical services there, reducing memory use by up to 90% while improving real-time stream throughput.

I started building high-traffic backend systems for HuffPost, from advertising platforms and content APIs to reliable news delivery. I like being the steady bassline: improving the systems that hold products together while jamming with ML, product, mobile, and platform teams.