BIFOLD Colloquium 02/2025

Secure collaborative analytics with MPC made practical

Abstract: “The performance of MPC-based approaches is so low that practical applicability is not in sight”. This is a review excerpt of a paper I co-authored, describing our vision to use MPC (multi-party computation) for secure data analytics in the cloud. In this talk, I will share how – 4+ years later –  we have realized this unlikely vision and more. I will first explain the legitimate skepticism of the particular reviewer and why past results indicated that MPC protocols were impractical for complex queries with joins and aggregations. I will then argue that careful system design and cross-layer optimizations can not only amortize MPC costs but also achieve scalability to large inputs and complex analytics, without compromising security. I will present the BU Secure Data Analytics Stack, our unified software architecture for secure collaborative data analytics in untrusted clouds. Finally, I will show performance results for secure relational and time series analytics on input sizes that a few years ago were only possible with hardware enclaves.