• Led a 6-person team to build Affirm’s first-generation Data Lake (Apache Iceberg), saving $2M/year, accelerating batch pipelines by 40%, and improving dataset availability by 67%.

• Migrated from AWS Redshift to Snowflake, cutting costs by 75% ($800K → $200K annually) and ensuring reliability ahead of peak BFCM (Black Friday / Cyber Monday) 2020 traffic.

• Designed a Spark Privacy Pipeline handling 2,500+ monthly data requests, preventing $2.5M/month in potential regulatory fines.

• Scaled the pipeline to process 500B+ events per run on Amazon EMR with 3,000 cores and 20,000+ tasks, ensuring reliability at scale.

• Solely architected an analytics (DBT) platform running 350+ datasets with CI/CD and monitoring, enabling 30+ analysts to set up in 10 minutes (vs. 2 days), automate weekly deployments (vs. 6 hours), and prevent SLA breaches from broken models.

• Taught Introduction to Relational Databases (CPSC 304).

• Built a Scala service to validate 400,000 SQL queries against bi-weekly PrestoDB upgrades to detect backwards compatibility breaks and performance regressions.

• Developed a full-stack application to visualize deployments, resources, and permissions; created metrics and alarms to minimize security risks.

• Shipped a macOS application to simplify authentication from 6 to 2 steps for data services.

• Created a CLI to manage deployments, enabling self-served services such as Presto, Mode, YARN, Azkaban for teams.