I work on data infrastructure at Automattic: the pipelines that move analytics events around, and the clusters and monitoring underneath them. I've been here since 2013, working remotely from Portland, OR.
These days I work on Parse.ly's realtime analytics pipeline, built on Kafka, Apache Storm, Redis, Cassandra and Elasticsearch, along with its Spark-based batch system. Before that I spent years on Automattic's Hadoop platform, built out Elasticsearch clusters for WordPress.com, and set up ClickHouse for WP Cloud.
Most of that code lives in private repositories, so what's public here is mostly side projects and upstream fixes.
Things I enjoy: patching up leaky pipelines, tracking down odd performance problems, and making things simpler, one red diff at a time.
Open source:
- heroku-wp: WordPress on Heroku.
- elasticsearch-whatson: an Elasticsearch cluster visualizer.
- pystorm: Python multi-lang support for Apache Storm.
- php-thrift-sql: a PHP client for Hive and Impala.
- Occasional patches to Apache Bigtop, Knox and Airflow.