Presto-Native Noisy Aggregations for Privacy-Preserving Workflows
summary
At Meta, Presto gains built-in DP primitives so SQL returns noisy aggregates instead of raw counts, enabling privacy across billions of users while keeping queries fast. Implementations include noisy approximate percentile, noisy count/sum/avg/distinct and min/max; some primitives are already in production and upstream open source PRs. The system pairs these with automatic query rewriting and privacy-preserving ETL pipelines so analysts send familiar SQL, pipelines and web maps run with enforced budgets, and teams see fewer deployment bugs and policy violations.
AI-written summary · about
related talks
22mIntroduction to Presto: The SQL Engine for Data Platform Teams
46mPresto - Today and Beyond - The Open Source SQL Engine for Querying all Data Lakes
22mPresto for Your Fast, Open SQL Lakehouse
32mData @Scale Seattle- Eric Hwang introduced by Ganapathy Krishnamoorthy
20mShare data without revealing personal information
27mPrivacy-preserving statistical analyses
45mMaking Privacy Technology Accessible: Benchmarks and Platforms
12mPlume: Differential Privacy at Scale