ZQL is a small domain specific language to express computations on private data. The ZQL compiler and tool chain turns these computations into efficient zero-knowedge protocols, to ensure the privacy and the integrity of the program execution. The programmer only has to provide a small source query, encoding business logic, and the ZQL compiler synthesizes automatically all the code for data sources to sign inputs; provers performing the computation and proofs; and verifiers checking the computations results provided are correct.

ZQL could be useful in a number of contexts:

• Privacy for billing, on-line profiling and analytics.
• Privacy friendly pay-as-you-drive insurence, pay as you go ticketing, or pay-per-view media streaming.
• On-line wallets or e-cash.

In all these settings customer information is certified by some meter or service, processed on clients and the results verified on a server. The raw readings do not have to be seen by the service, yet the clients cannot cheat about the result!

Presentations and Publications related to ZQL:

• Cedric Fournet, Markulf Kohlweiss, George Danezis and Zhengqin Luo. ZQL: A Compiler for Privacy-Preserving Data Processing. MSR-TR-2013-25. February 2013.
• George Danezis. ZQL presentation at EPFL. December 2012.
• Cedric Fournet. ZQL presentation. February 2013.

Related papers, on secure client side programming:

• George Danezis, Markulf Kohlweiss, Benjamin Livshits, Alfredo Rial.Private Client-Side Profiling with Random Forests and Hidden Markov Models. Privacy Enhancing Technologies 2012, Vigo Spain.
• George Danezis and Benjamin Livshits. Towards Ensuring Client-Side Computational Integrity (Position Paper). The ACM Cloud Computing Security Workshop. CCSW 2011, Chicago, USA, October 17, 20011.
• Alfredo Rial and George Danezis. Privacy-Preserving Smart Metering. Proceedings of the 2011 ACM Workshop on Privacy in the Electronic Society, WPES 2011, Chicago, USA, October 17, 20011.