SmartNoise

A joint venture between Microsoft and Harvard’s OpenDP to provide production-ready differential privacy for sensitive datasets.

SmartNoise implements differential privacy through a robust SDK and system service (OpenDP). It allows analysts to query private databases while injecting calibrated noise to prevent individual re-identification. The toolkit handles SQL-to-DP translation and supports advanced mechanisms like the Laplace and Gaussian noise models. Organizations use it to share insights from medical records or financial transactions without exposing raw PII (Personally Identifiable Information).

https://github.com/opendp/smartnoise-sdk