I make distributed data systems faster at petabyte scale.
Architect on the Data Processing Engines team at AWS - Spark, Iceberg, and the open-source surface of EMR.
- 🛠️ Maintainer of the official AWS EMR open-source repos (aws-emr-utilities, aws-emr-best-practices) one of fewer maintainers across 1000s of companies running Amazon EMR.
- 📈 Led Spark performance at PB scale 45% runtime efficiency gain and 60% shuffle-spill reduction across 100+ customers.
- 📘 Author of the EMR Spark Best Practices Guide and Top 10 Best Practices for Amazon EMR Serverless on the AWS Big Data Blog.
- 🎤 4x AWS re:Invent speaker — most recently ANT339 on GoDaddy's EMR Serverless migration: 62% cost reduction, 50% faster execution, 5x faster dev cycles.
- 🔥 Contributing to Apache Spark, Velox, and Karpenter.
- 🔬 Published researcher MS in Artificial Intelligence.
🚧 Currently: contributing bandwidth-weighting support to Karpenter; advancing GenAI-assisted diagnostics for Spark on EMR.
Interests: Shuffle optimization, query engines, Iceberg lineage, and petabyte-scale data processing.
Previously: 18+ years across AWS, DISH Network, and EchoStar — cloud-native data platforms, on-premise EDW migration, Wall Street regulatory reporting.
📫 Reach me on LinkedIn.